ixon Driving the absolute best from EMCCD technology Including the ixon Ultra Megapixel 26 fps (full frame) > 90% QE USB 3.0

Transcription

1 ixon Driving the absolute best from EMCCD technology Including the ixon Ultra Megapixel 26 fps (full frame) > 90% QE USB 3.0

2 ixon The Industry s Highest Performance Scientific EMCCD Cameras Andor Technology pioneered the world s first scientific Electron Multiplying CCD (EMCCD) cameras, shipping the initial cameras back in 2000 and winning the Photonics Circle of Excellence award. At that time, Andor coined the name Electron Multiplying CCD (EMCCD), which has been adopted right across this burgeoning industry. Since then, Andor has consistently set progressively higher EMCCD performance standards with our successive ixon series of deep-cooled, vacuum sealed, quantitative EMCCD cameras. For example, Andor introduced the first back-illuminated EMCCDs in January 2002, alongside our unique Baseline Clamp solution for enhanced quantitative performance. Andor s method for achieving industry-lowest Clock Induced Charge (CIC) was introduced in early 2003 and our benchmark quantitative and linearized EM gain control (RealGain ) and patented EM gain recalibration technology (EMCAL ) was innovated in January In 2010, Andor introduced the ixon3 series that brought a number of customer requested features, such as one-click application optimization (OptAcquire) and the ability to calibrate data in either photoelectrons or photons (Count Convert). The ixon Ultra has built upon this rich performance and feature set, raising the bar markedly higher still by driving frame rates up to 3x more than our already industry leading speed performance. This opens new possibilities for the majority of EMCCDenabled applications that benefit from single photon sensitivity at fast frame rates Andor shipped first EMCCD cameras Pioneering EMCCD cameras Photonics Circle of Excellence Award 2001 ixon Flagship high-end EMCCD camera platform First to introduce back-illuminated sensors Deepest cooling and lowest noise Fastest frame rates 2006 ixon+ New innovations and greater sensitivity Patented EMCAL and RealGain TM World s first Megapixel back-illuminated EMCCD (888 model) Maintaining market leader status The EMCCD advantage Current trends in photon measurement are placing unprecedented demands on detector technology to perform at significantly higher levels of sensitivity and speed. Electron Multiplying CCD (EMCCD) technology has been designed to respond to this growing need, unlocking new and innovative experimental prospects. EMCCDs operate by amplification of weak signal events (down to single photons) to a signal level that is well clear of the read noise floor of the camera at any readout speed. Importantly, this on-chip amplification process is realized without sacrificing the photon collection capability of the sensor, with back-illuminated sensors offering up to 95% Quantum Efficiency (QE). ixon - The Microscopist s Choice In applications such as single molecule microscopy, super-resolution, live cell microscopy (including confocal), calcium signaling, transport/motile imaging and intracellular bioluminescence, weak, rapidly changing fluorescent signals from cells must be dynamically imaged. Andor s ixon technology offers an ideal detection solution. Ultra-sensitive detection capability in fluorescence microscopy facilitates use of lower excitation powers (thereby reducing photobleaching and phototoxicity) and lower dye concentrations. Since its pioneering introduction in 2000, Andor s EMCCD technology has been widely and highly successfully employed by microscopists throughout the world, resulting in an outstanding level of representation in high-profile publications ixon3 New customer requested functionality Superior ease of optimization Quantify data in electrons or photons Enhancements in noise, speed and quantitative stability ixon Ultra >60% faster x 56 fps USB universal Plug and Play Lower noise conventional CCD mode On the fly processing - direct output of raw data New EX2 Technology - extended EMCCD sensitivity ixon Ultra 888 Largest Field of View, 3x faster! 26 fps from 1024 x 1024; 93 fps from 512 x 512 First USB 3.0 enabled EMCCD With the ixon series, Andor have delivered a dedicated, truly high-end, ultra-sensitive scientific camera platform, designed specifically to extract the absolute best from Electron Multiplying CCD (EMCCD) technology across all critical performance specs and parameters. ixon - The Physicist s Choice The unique high-performance specifications of the optimized ixon range have been serving the physical scientist and astronomer in scenarios that demand more than simply an EM sensor in a camera. Andor has worked with numerous scientists to deliver solutions that work for their particular application requirements, such as providing effective charge purging immediately prior to acquisition, specific coatings, coupling to fiber optic scintillators and also specific interface requirements. As such the Andor ixon brand has been prevalent across a variety of demanding applications, such as photon counting, lucky astronomy, adaptive optics, Bose Einstein condensation (BEC) / ion trapping, single molecule detection / nanotechnology, neutron tomography, X-Ray/Gamma tomography, plasma diagnostics, Raman detection and thermoluminescence detection. 2 3

3 Why choose Andor s ixon high performance EMCCD? The principal reason for making use of Andor s ixon EMCCD technology is to ensure the absolute highest sensitivity from a quantitative scientific digital camera, particularly under fast frame rate conditions. In particular, truly exceptional speed performance is now available through the new ixon Ultra model. Andor s proven UltraVac vacuum technology, carrying a seven year warranty, is critical to ensure both -100ºC deep cooling and complete protection of the sensor. The ixon series of cameras are designed to be the most flexible yet easy to use EMCCDs on the market, optimizable for a wide variety of application requirements in a single click via the new OptAcquire feature. Furthermore, signal can be quantitatively calibrated in units of electrons or photons, either in real time or post-processing. Patented, pioneering technology offers automated recalibration of EM gain, alongside anti-ageing protection. Crucially, the ixon brand carries an outstanding reputation within the industry for quality and reliability, brandishing an unparalleled track record of minimal field failures. Andor s ixon Ultra and ixon3 EMCCD cameras deliver best performance across these core areas: Fastest Frame Rates ixon Ultra delivers up to 3x faster frame rates. Unique Optically Centered Crop Mode solution for additional frame rate boost (e.g. 569 fps with 128 x 128 ROI). Highest Sensitivity EMCCD Deepest cooling and lowest clock induced charge parameters complement the high QE performance of back-illumination. New EX2 Technology offers extended EMCCD sensitivity. Ultimate Longevity Essential vacuum protection of back-illuminated sensor meaning no QE degradation. Anti- Ageing technology and EMCAL, a pioneering auto-recalibration of EM gain. Unparalleled long-term performance capability. User Friendliness ixon remains the most versatile EMCCD camera available. Now, via OptAcquire, it can be optimized for different application requirements with a single click. RealGain - Select absolute EM gain direct from a linear and directly quantitative software gain scale. The EM gain you ask for is the EM gain you get. Quantitative: Baseline (bias) Clamp and Stabilized EM Gain all that should vary is your signal. Now with option to display signal in electrons or photons, real time or post convert. Flexibility Ability to fine tune performance for a wide range of application scenarios. 2 in 1 EMCCD and low noise CCD amplifiers (model dependent). Ability to turn off EM gain completely for rapid imaging under brighter conditions. Advanced Performance Through Software icam encompasses a set of unique innovations that empower ixon cameras to operate with market-leading exposure switching through imaging software packages. Direct Data Access ixon Ultra offers a unique Camera Link output port to facilitate direct access to raw data for on the fly processing, ideal for applications such as adaptive optics. Superior Acquisition Efficiency One Click Optimization Quantitative Stability Outstanding Quality Vacuum Longevity USB or PCI Lowest Noise Patented EM Gain Calibration Quantify in Electrons or Photons Fastest Frame Rates Direct Data Access -100ºC Cooling 4 5

4 ixon Range The ixon portfolio encompasses a number of model variations, offering solutions for a wide range of application requirements. Whether your needs are guided more by speed, resolution, field of view, wavelength dependence or simply budget, the ixon series of market leading EMCCD cameras will provide a match. Model ixon Ultra 888 Core Attributes Field of View, Sensitivity and Speed Sensor Format QE Options Pixel Size 1024 x 1024 BV, EX2, UVB 13 µm ixon Ultra 897 Sensitivity and Speed 512 x 512 BV, EX2, UVB 16 µm ixon3 860 Dedicated to Speed; Large Pixel Frame Rate 128 x 128 BV, UVB 24 µm 515 fps 26 fps (670 fps with 128 x 128 Crop Mode) 56 fps (595 fps with 128 x 128 Crop Mode) Interface USB 3.0 USB 2.0 PCI- Express Quantum Efficiency (%) UVB EX2 BV, BVF EX Wavelength (nm) Resin-dentine interface created using an etching-and-rinse bonding agent previously mixed with 0.1% rhodamine-b and subsequently applied in deep dentine to aid the visualization of resin tags, adhesive and interdiffusion layers. A reflective 10% silver nitrate solution was used to evaluate the microporosities within the hybrid layer of the resin-dentine interfaces. The silver grains located within the microporosities of the interdiffusion layer have diameters of <1 µm. Image courtesy of Salvatore Sauro, King s College London Dental Institute 6 7

5 Features and Benefits ixon Ultra 888 Field of View and Sensitivity... now 3x faster! The highly innovative ixon Ultra 888 Megapixel, back-illuminated EMCCD camera, offers single photon sensitivity across a large field of view, at 25 fps. Building on a rich history of first to market innovation, the supercharged ixon Ultra 888, represents a massive performance boost for the largest available EMCCD sensor, as well as the first USB 3.0 enabled EMCCD camera. The ixon Ultra 888 has been fundamentally re-engineered to facilitate 3x overclocking of the pixel readout speed to an unprecedented 30 MHz, whilst maintaining quantitative stability, thrusting the full frame rate performance to video rate. Furthermore, Andor s unique Crop Mode can be employed to further boost frame rates from a user defined sub-region, for example pushing a 512 x 512 sub-array to 93 fps and a 128 x 128 area to 670 fps. With a 1024 x 1024 sensor format and 13 µm pixel size, the resolving power, field of view and unparalleled speed of the ixon Ultra 888 render it the most attractive and versatile EMCCD option for demanding applications. These include single molecule detection, super-resolution microscopy, live cell imaging and high time resolution astronomy. Key Applications Single Molecule Detection Super Resolution (PALM, STORM) TIRF Microscopy Selective / Single Plane Illumination Microscopy (SPIM) Vesicle Trafficking Spinning Disk Confocal Microscopy Ion Signaling (Calcium Flux) Voltage Sensitive Dyes Cell Motility Key Specifications Whole Genome Sequencing FRET / FRAP Fluorescence Correlation Microscopy (Multi-beam) Microspectroscopy / Hyperspectral imaging Lucky Astronomy Adaptive Optics Single photon counting Overclocked to 30 MHz readout 13.3 x 13.3 mm sensor Largest field of view EMCCD available. Optically Centred Crop Mode (live cell super-resolution) First USB 3.0 enabled EMCCD EX2 Technology (optional) TE cooling to -90 C RealGain Fringe Suppression (optional) OptAcquire Count Convert EMCAL icam Minimal Clock-Induced Charge UltraVac Spurious Noise Filter Direct Data Access Enhanced photon counting Modes Superior Baseline Clamp and EM Stability Lower Noise CCD Mode Pushes frames to 26 fps (full frame); 93 fps with 512 x 512 Crop Sensor Mode. Continuous imaging with fastest possible frame rate from centrally positioned ROIs. Highly enabling for live cell super-resolution and much more (e.g. 251 fps with 256 x 256 ROI). Enhanced bandwidth and simple connectivity. Extended QE response, beyond standard back-illuminated. Elimination of dark current detection limit. Absolute EMCCD gain selectable directly from a linear and quantitative scale. Reduced etaloning in NIR. Optimize the highly flexible ixon for different application requirements at the click of a button. Quantitatively capture and view data in electrons or incident photons. Applied either in real time or postprocessing, Count Convert does this important conversion for you. Patented user-initiated self-recalibration of EM gain. Exposure time fast switching provides market leading acquisition efficiency. Unique pixel clocking parameters, yielding minimized spurious noise floor. Critical for sustained vacuum integrity and to maintain unequalled cooling and QE performance, year after year. Seven year vacuum warranty. Intelligent algorithms to filter clock induced charge events from the background. Real time or post-processing. Camera Link output port to facilitate direct access to data for on the fly processing. Intuitive single photon counting modes to overcome multiplicative noise. Real time or post-processing. Essential for quantitative accuracy of dynamic measurements. 2 in 1 flexibility. EMCCD for ultra-sensitivity at speed, conventional CCD for longer acquisitions. Active Pixels 1024 x 1024 FPGA Timestamp Hardware generated timestamp with 10 ns accuracy. Pixel Size (W x H; μm) 13 x 13 Image Area (mm) 13.3 x 13.3 Active Area Pixel Well Depth (e - ) 80,000 Max Readout Rate (MHz) 30 Frame Rates (frames per sec) 26 (full frame) - 9,610 Read Noise (e - ) < 1 with EM gain QE max > 90% (EX2 available) Additional features of the ixon Ultra 888 include high bandwidth USB 3.0 connectivity, a lower noise CCD mode and an additional Camera Link output. This offers a unique ability to directly intercept data for on the fly processing, ideally suited to applications such as adaptive optics. Simultaneously, the ixon Ultra maintains all the advanced performance attributes and a rich customer requested feature set that have defined the ixon range to date, such as deep vacuum cooling to -95ºC, extremely low spurious noise and EM Gain calibration. Count Convert functionality allows real time data acquisition in units of electrons or incident photons and OptAcquire facilitates one-click optimization of this versatile camera to a variety of application conditions Field of View Comparison between ixon Ultra models. The 888 model has a x2.6 greater sensitive area than the 897 model. 8 9

6 Features and Benefits ixon Ultra 897 Overclocked to 17 MHz readout Optically Centered Crop Mode (Live cell super-resolution mode) Pushes frames to 56 fps (full frame); 595 fps with 128 x 128 cropped sensor mode. Continuous imaging with fastest possible frame rate from centrally positioned ROIs; 569 fps with 128 x 128 ROI. Highly enabling for live cell super-resolution and much, much more. Ultimate Sensitivity... Supercharged! Facilitated by a fundamental redesign, the ixon Ultra 897 takes the popular backilluminated 512 x 512 frame transfer sensor and overclocks readout to 17 MHz, pushing speed performance to an outstanding 56 fps (full frame), whilst maintaining quantitative stability throughout. Ultimate sensitivity is also attained through deep thermoelectric cooling down to -100 C and industry-lowest clock induced charge noise. New EX2 Technology offers extended QE performance. EX2 Technology TE cooling to -100 C Fringe Suppression RealGain OptAcquire Count Convert EMCAL Extended QE response. Critical for elimination of dark current detection limit. Reduced etaloning in NIR. Absolute EMCCD gain selectable directly from a linear and quantitative scale. Optimize the highly flexible ixon for different application requirements at the click of a button. Quantitatively capture and view data in electrons or incident photons. Applied either in real time or post-processing, Count Convert does this important conversion for you. Patented user-initiated self-recalibration of EM gain. Additional unique features of the ixon Ultra include Optically Centered Crop Mode for superb speed from ROIs and direct raw data access for on the fly processing. EMCCD and conventional CCD readout modes provide heightened application flexibility, with a new low and slow noise performance in CCD mode. The significant speed boost offered in the ixon Ultra 897 facilitates a new level of temporal resolution to be attained. This is ideal for speed challenged low-light applications such as super-resolution microscopy, single molecule tracking, ion signaling, cell motility, single photon counting, lucky astronomy and adaptive optics. The extremely low noise of the ixon 897 coupled with the new overclocked speed performance will place this model at the forefront of consideration when it comes to upgrading the high end imaging performance of your laboratory. Key Specifications Active Pixels 512 x 512 Pixel Size (W x H; μm) 16 x 16 Image Area (mm) 8.2 x 8.2 Active Area Pixel Well Depth (e - ) 180,000 Max Readout Rate (MHz) 17 Frame Rates (frames per sec) 56-11,074 Read Noise (e - ) < 1 with EM gain QE max > 90% (EX2 available) icam Minimal Clock-Induced Charge USB 2.0 UltraVac Spurious Noise Filter Direct Data Access Enhanced photon counting Modes Superior Baseline Clamp and EM Stability Lower Noise CCD Mode Exposure time fast switching provides market leading acquisition efficiency. Unique pixel clocking parameters, yielding minimized spurious noise floor. True universal plug and play. No internal card required for compatibility. Operates readily on laptop at top speed. Critical for sustained vacuum integrity and to maintain unequalled cooling and QE performance, year after year. 7 year vacuum warranty. Intelligent algorithms to filter clock induced charge events from the background. Real time or post-processing. Camera Link output port to facilitate direct access to data for on the fly processing. Intuitive single photon counting modes to overcome multiplicative noise. Real time or post-processing. Essential for quantitative accuracy of dynamic measurements. 2 in 1 flexibility. EMCCD for ultra-sensitivity at speed, conventional CCD for longer acquisitions. FPGA Timestamp Hardware generated timestamp with 10 ns accuracy. Key Applications Market leading back-illuminated EMCCD, now over 60% faster Single Molecule Detection Super Resolution (PALM, STORM) TIRF Microscopy Selective / Single Plane Illumination Microscopy (SPIM) Vesicle Trafficking Spinning Disk Confocal Microscopy Ion Signaling (Calcium Flux) Voltage Sensitive Dyes Cell Motility Whole Genome Sequencing FRET / FRAP Fluorescence Correlation Microscopy (Multi-beam) Microspectroscopy / Hyperspectral imaging Lucky Astronomy Adaptive Optics Single photon counting 10 11

7 Features and Benefits ixon3 860 Lightning Speed and Sensitivity Andor s ixon3 860 back-illuminated EMCCD is designed for dedicated rapid imaging of low light events, without use of Region of Interest (ROI). The 128 x 128 sensor of the ixon3 860 combines > 500 fps frame rate with single photon detection capability and > 90% Quantum Efficiency. Thermoelectric cooling down to -100 C minimizes EM-amplified dark current, whereas industry fastest vertical shift speeds minimize both clock induced charge noise and vertical smear during frame transfer. The absolute EM gain multiplication can be varied linearly from unity up to a thousand times directly via RealGain TM, a true quantitative EM gain scale. Sub-millisecond biology is readily accessible through use of sub-array selection and pixel binning. The speed and sensitivity of the 860 also renders it ideal for adaptive optics. 513 full fps Fast frame rates ideal for ion signaling microscopy and adaptive optics. TE cooling to -100 C RealGain OptAcquire Count Convert icam UltraVac Crop Mode Spurious Noise Filter Enhanced photon counting Modes Critical for elimination of dark current detection limit. Absolute EMCCD gain selectable directly from a linear and quantitative scale. Optimize the highly flexible ixon3 for different application requirements at the click of a button. Quantitatively capture and view data in electrons or incident photons. Applied either in real time or post-processing, Count Convert does this important conversion for you. Exposure time fast switching provides market leading acquisition efficiency. Critical for sustained vacuum integrity and to maintain unequalled cooling and QE performance, year after year. Specialized acquisition mode for continuous imaging with fastest possible temporal resolution. Intelligent algorithms to filter clock induced charge events from the background. Real time or post-processing. Intuitive single photon counting modes to overcome multiplicative noise. Real time or post-processing. Superior Baseline Clamp and EM stability Essential for quantitative accuracy of dynamic measurements. Key Specifications Key Applications Active Pixels 128 x 128 Single Molecule Detection Pixel Size (W x H; μm) 24 x 24 Calcium Flux Image Area (mm) 3.1 x 3.1 Voltage Sensitive Dyes Active Area Pixel Well Depth (e - ) 160,000 Adaptive Optics Max Readout Rate (MHz) 10 Frame Rates (frames per sec) Read Noise (e - ) QE max > 90% 513 (full frame) up to several thousands 10 MHz < 1 with EM gain FRET Fluorescence Correlation Spectroscopy (FCS) Nanopore array The use of the highly sensitive and ultra-fast back-illuminated ixon 860 EMCCD is central to our high-speed single molecule gene sequencing method, as we rely on fast multi-color optical readout from many nanopores simultaneously. EMCCD Prof Amit Meller, Associate Professor of Biomedical Engineering and Physics, Boston University Schematic diagram of the single molecule Optipore sequencing method, with an illustration of the enzyme-free nanopore array through which DNA strands are electrophoretically drawn. ixon3 860 cameras were used to capture two color sequencing data at 1000 fps. Courtesy of Prof. Amit Meller, Dept. of Biomedical Engineering and Physics, Boston University 12 13

8 ixon Performance and Innovations Industry Fastest Frame Rate Maximum frame rate performance in EMCCDs is a function of two parameters; (1) Pixel Readout Speed (horizontal); (2) Vertical Clock speed. The former dictates how rapidly charge is pushed horizontally through the EM gain register and the remaining readout electronics, while the latter dictates the speed at which charge is vertically shifted down through both the exposed sensor area and masked frame transfer area of the chip. ixon offers industry fastest vertical shift speeds, resulting in faster frame rates and reduced smearing, significantly faster under commonly employed conditions of sub-array/binning. Notably, the ixon Ultra 897 overclocks the pixel readout speed from the standard 10 MHz to 17 MHz, further boosting the frame rate by >60%, yielding 56 fps (full frame). The new ixon Ultra 888 takes this a big step further, thrusting clock speed to 30 MHz! This permits video rate frame rate from this large field of view sensor, and enables as fast as 93 fps from a 512 x 512 ROI in Crop Mode. Key Features ixon Ultra overclocked up to 30MHz pixel readout speed: 3x faster full frame rate Fastest vertical shift speeds yield further speed gains with ROI / binning Pushing Frame Rates with Crop Mode Through this acquisition mode, significant increases in frame rates are accomplished by fooling the sensor into thinking it is smaller than it actually is. In standard sub-array/ ROI readout mode each frame still carries the time overhead to readout all pixels to the left and right of the selected area and to vertically shift all pixels above and below the selected area. The charge from these pixels is then dumped before an image is sent from camera to PC. In Crop Mode, the number of pixel readout steps outside of that required to read out the requested sub-array is significantly reduced, resulting in markedly higher frame rates. However, this mode requires that light is not allowed to fall onto the area of the sensor outside of the defined active sub-area. In optical microscopy, this can be realized in conjunction with the new OptoMask accessory, which inserts easily between the microscope output and the camera. Using the OptoMask, a sub-array can be readily defined through positioning of the masking blades, and a cropped area matched to this in software. Crop Mode The active imaging area of the sensor is defined in a way that only a small section of the entire chip is used for imaging. The remaining area has to be optically masked to prevent light leakage and charge spill-over that would compromise the signal from the imaging area. By cropping the sensor, one achieves faster frame rates because the temporal resolution will be dictated only by the time it requires to read out the small section of the sensor. Minimized smearing through faster vertical shifts Optically Centered Crop Mode for industry fastest ROI speeds, ideal for live cell super-resolution microscopy Optically Centered Crop Mode: Enabling live cell super-resolution See page 32 for Maximizing frame rate technical note The ixon Ultra now comes with Optically Centered Crop Mode, which gives the user the option to break away from the corner tethered requirement of standard crop mode and select a number of pre-defined ROIs that are located in the centre of the image field. This is achieved with only minimal sacrifice in achievable frame rate, for example a 128 x 128 optically centered ROI delivering 697 fps. Optically centring of the ROI makes this mode extremely appealing to a number of microscopy techniques, including pointillism live cell super-resolution microscopy. For example, the camera can be operated in full 512 x 512 resolution at a frame rate suited to generation of fixed cell superresolved images, then Optically Centered Crop Mode can be invoked with a 128 x 128 ROI for generation of super-resolved live cell images showing dynamic events. Array size Binning 512 x x x x x x x 1 1 x 1 93 (78) 190 (251) 670 (697) 2053 (1319) x (143) 350 (426) 1150 (1019) 3123 (1646) x (245) 601 (653) 1772 (1504) 4109 (1857) ixon Ultra with OptoMask enabling faster frame rates in Crop Mode Frame rates achievable by the ixon Ultra 888 in Crop Mode - Optically Centered Crop Mode frame rates in brackets The ixon Ultra 888 can achieve a blistering 251 fps from a 256 x 256 ROI in Optically Centered Crop Mode 14 15

9 ixon Performance and Innovations EX2 Technology Extended QE from Dual AR sensor coating Selected ixon models are now available with a new Dual Anti-Reflection coating applied to the back-illuminated sensor, affording a significant enhancement of the Quantum Efficiency performance. Fringe Suppression Sensors Selected ixon models are now available with a new Fringe Suppression property in the sensor design, reducing spatial etaloning effects that can arise through monochromatic imaging in the Near Infra-Red (NIR) wavelength range. Available on the new speed-boosted ixon Ultra 897 and 888 models, EX2 technology facilitates broadening of the QE range of the back-illuminated sensors through implementation of a new dual AR coating process, developed by sensor manufacturer e2v. The net effect is to offer significantly improved sensitivity in both the blue and NIR wavelength regions, whilst maintaining ~ 90% QE across the remainder of the visible region 100 See page 36 for Extended QE and Fringe Suppression technical note Etaloning is particular to back-illuminated sensors and is caused by interference between reflections off the front and back parallel sensor surfaces. For NIR applications, such as imaging of Bose Einstein Condenates, etaloning effects, often observed more notably beyond ~ 750nm, can sometimes restrict the ability to perform high integrity quantitative imaging. Fringe Suppression refers to a sensor design modification that significantly reduces the amplitude of etaloning effects. The design has been implemented by sensor manufacturers e2v and based on a tried and trusted process, successfully validated over several years of CCD manufacture. Fringe Suppression is available on ixon Ultra 897 and 888 models. Quantum efficiency (%) C EX2 BV We have successfully used Andor ixon cameras for many years for super-resolution microscopy, the resolution and sensitivity of these cameras is exceptional. Dr. Mike Heilemann, Institute of Physical and Theoretical Chemistry, Goethe- University, Germany Wavelength (nm) C Quantum efficiency (%) Wavelength (nm) EX2 BV Back-illuminated EMCCD sensor QE curves, comparing standard BV mid-band AR coating versus new EX2 dual AR coating. Reference data available at +20 C and -100 C sensor cooling temperatures

10 ixon Performance and Innovations OptAcquire Flexibility need not be complicated The control architecture of the ixon is extremely flexible, meaning the camera can be adapted and optimized for a wide variety of quantitative experimental requirements, ranging from single photon counting through to slower scan, 16-bit Dynamic Range measurements. However, we are starkly aware that optimizing EMCCD technology is far from trivial, with various set-up parameters influencing and trading off between different camera performance characteristics. We have developed OptAcquire, a unique interface allowing users to conveniently choose from a predetermined list of camera set-up configurations. The user need only choose how they would like their camera to be optimized, e.g. for Sensitivity and Speed, Dynamic Range and Speed, Time Lapse. Parameters such as EM gain value, vertical shift speed, vertical clock amplitude, pre-amp sensitivity and horizontal readout speed will then be optimized accordingly, behind the scenes. Furthermore, the option exists to create additional user-defined configurations. Key Features Convenient one-click set-up Opens the market leading flexibility of the ixon to less advanced users Optimize for range of experimental requirements Create additional user defined modes Count Convert ixon offers the capability to quantitatively capture and present data in units of electrons or photons, this important conversion is applied either in real time or as a post-conversion step. The standard way to present quantitative data in scientific detectors has been in units of counts, relating to the digitized steps of the Analogue to Digital Converter (ADC) used in the camera. Each Analogue to Digital Unit (ADU) relates to a precise number of photo-electrons that were generated originally from photons striking and being captured by the detector pixel. In the ixon, this conversion factor is very accurately recorded within the camera. Knowing this value, alongside the EM gain (RealGain TM ) and baseline (bias) offset, facilitates back calculation from the signal in ADU counts per pixel to the signal in electrons per pixel. Furthermore, knowledge of the Quantum Efficiency (QE) and light throughput properties of the camera at each wavelength enables this process to be taken a step further, allowing the signal to be estimated in photons incident at each pixel, provided the spectral spread of the signal is not too broad. The Count Convert functionality of the ixon provides the flexibility to acquire data in either electrons or incident photons, with negligible slow down in display rate. Furthermore, the option exists to record the original data in counts and perform this important conversion to either electrons or photons as a post-conversion step, while retaining the original data. Key Features Quantify data in electrons or incident photons Convenient estimate of sample signal intensity at the detector Real time or post-convert Reference between different samples, users and set-ups Meaningful signal relating to PALM/STORM localization accuracy See page 40 for OptAcquire technical note We have been using the EMCCD range of cameras from Andor since their launch in Our original 887 camera is still fully functional more than 10 years later. These cameras have revolutionized the ability to resolve Ca 2+ signaling mechanisms in isolated cells. The ixon range of cameras have been incredibly reliable in our lab and have facilitated rapid, low noise, imaging at relatively low laser intensities. See page 35 for Count Convert technical note Dr. Mark Hollywood, Dundalk Institute of Technology, Ireland 18 19

11 ixon Performance and Innovations Deep Thermoelectric Cooling Single thermal electrons are amplified by the EMCCD gain mechanism. Deep vacuum TE cooling is critical to optimize the sensitivity performance of EMCCD sensors, otherwise the raw sensitivity will be compromised, even under conditions of short exposures Key Features Cooling down to -100ºC Lowest EM-amplified dark current Spurious Noise Filter It can still be desirable to optionally filter the remaining spurious noise (Clock-Induced Charge or photons) to give as black a background as possible, eradicating any remaining salt and pepper noise. It is important to utilize noise selection and filter algorithms that are intelligent enough to accomplish this task without impacting the integrity of the signal itself. This is realized through the new Spurious Noise Filter (SNF) functionality of ixon, which offers the user a choice of advanced algorithms to try. SNF can be applied either in real time or as a post-processing step. Fewer pixel blemishes (hot pixels) Low power consumption vacuum cooling Vertical Shift Speed = 0.5µs Before SNF After SNF -70ºC -95ºC Line intensity profiles across a row of 512 pixels of the 897 model, taken from DARK images at three different cooling temperatures, 29 ms exposure times. The cleanest noise floor is clearly seen under conditions of deep cooling, even for such short exposure times. Before and after application of the ixon Spurious Noise Filter Images of extremely weak LED signal (signal intensity typical of weak luminescence experiments) acquired with ixon3 888 at cooling temperatures -70ºC and -95ºC (water cooling to achieve latter), 120 sec exposure times, sub-region show. The need to push to such deeper cooling temperatures can be readily observed under such extreme low light conditions. Minimized Clock-Induced Charge After having minimized dark current through deep cooling, the remaining detection limit in back-illuminated EMCCDs is given by the number of Clock-Induced Charge noise events. Andor s industry-exclusive combination of high resolution clocking parameters and sub-microsecond clock speeds are fundamental to minimizing CIC, enabling truly high-end EMCCD sensitivity to be claimed. 3.3µs 1.7µs 0.5µs DARK IMAGES taken with the ixon3 897 at x1000 gain at different vertical shift speeds, 29 ms exposure time. Cooling temperature was C to ensure minimal dark current contribution

12 Superior Quantitative Stability Baseline Clamp ON ixon Performance and Innovations UltraVac Permanent Vacuum Head It is important that a back-illuminated sensor is housed in a hermetically sealed permanent vacuum head with minimized outgassing, otherwise both cooling performance and the sensor QE will steadily degrade. It is this compelling reason that drove Andor to develop UltraVac. Andor s proprietary UltraVac process has a proven track record of field reliability, accumulated over more than 15 years of shipping high-end vacuum cameras. UltraVac also enables use of only one input window, improving photon-throughput by 8%. Key Features No QE degradation Sustained deep TE cooling No maintenance / re-pumping One input window No condensation Vacuum Window (Anti-reflection Coated) Permanent, All-Metal Seal Heat Removal The ixon is well regulated in terms of both Baseline (bias offset) rigidity and superior EM gain stability, lending for enhanced quantitative reliability throughout and between measurements. EM gain Stability Mean ROI signal / ADU EM Gain stability in the ixon Ultra 55 fps. 500 frame kinetic series; frame transfer (overlapped) acquisition; 17.8 ms exposure time; x300 EM Gain. RealGain TM, Anti-Ageing and EMCAL Baseline Clamp OFF ixon Baseline Clamp (bias stability) in operation In early 2006, Andor once again raised the bar by introducing some significant new technology innovations. These particular pioneering steps, were to set new high standards in quantitative EMCCD usage and general EMCCD longevity expectations, which others in the industry are now adopting. Mean ROI signal / ADU Mean ROI signal / ADU CCD TE Cooler Schematic of a Permanent Vacuum Head 7 Year Vacuum Warranty RealGain Select absolute EM gain direct from a linear and directly quantitative software scale, x1 to x1000. The EM gain you ask for is the EM gain you get. Anti-Ageing Internally configured to significantly inhibit saturation-induced decay of EM gain. EMCAL - Innovative user-initiated self-recalibration of EM gain, utilizing a patented method of automated EM gain assessment and Andor s unique Linear and Real Quantitative gain implementation. Temperature Compensated - Calibration holds across all cooling temperatures. No need to recalibrate on each use in multi-user laboratories and facilities. Unlike other vacuum EMCCDs on the market, the ixon family has now been shipping with a vacuum enclosed sensor for almost 10 years, with statistical data that substantiates our extremely robust vacuum claims. With the ixon, Andor are proud to offer an extended 7 year warranty on the vacuum enclosure as standard. ANDOR S EXTENDED YEAR WARRANTY Enhanced photon counting To successfully photon count with EMCCDs, there has to be a significantly higher probability of seeing a photon spike than seeing a dark current/cic noise spike. The ixon Ultra 897 combines deepest thermoelectric cooling and low CIC performance, yielding market leading photon counting performance and higher contrast images. photon counting Higher CIC photon counting Lowest CIC Real-time and post process photon counting The advanced photon counting modes of the ixon allow for both real time and post-process photon counting. The latter offers the flexibility to trial and error photon count a pre-recorded kinetic series, trading-off temporal resolution vs SNR

13 ixon Performance and Innovations Direct Data Access (ixon Ultra) Under standard operation the ixon Ultra uses the USB interface (USB 2.0 or USB 3.0, model dependent) for all control and data transfer with the PC. However, some users require a more direct access to the image data stream, in order that they can perform real-time analysis, possibility using external hardware. Such operation can be particularly important for rapid closed feedback applications such as adaptive optics. Direct real time access to data can also be useful for data intensive applications such as super-resolution microscopy or whole genome sequencing, whereby it can be desirable to carry out real time processing of data on an external GPU, for example. In order to facilitate such functionality, the ixon Ultra includes an additional Camera Link output port. The Camera Link channel intercepts the image data stream in the camera head immediately after the on-head FPGA processing step, but before the USB frame buffer, therefore undergoes the same amount of on-head image processing. 2 in 1 Performance - EMCCD and CCD Three of the back-illuminated ixon models offer 2 in 1 performance flexibility, in terms of operating as a single photon EMCCD or a low noise conventional CCD, readily user selectable through software selection. Such versatility is attractive in laboratories that can require the camera to operate in low light conditions under both fast and slow frame rates. In photon starved applications, choosing the EMCCD amplifier usually yields better signal to noise ratio when under faster frame rates conditions (> 1 fps), whereas often the CCD amplifier can yield better signal to noise ratio when longer exposures can be applied and when the sensor can be read out slowly (i.e. seconds per frame rather than frames per second ). This a rule of thumb guide however, and often the choice of amplifier depends ultimately on the light levels available during desired exposure time. Usually it is worth experimenting with the CCD amplifier if the temporal demands are sufficient to readout the sensor at 1 MHz or slower. Key Features Direct Data Access via Camera Link output Minimal latency or jitter USB data stream concurrently accessible Compatible with any Camera Link card interface On-the-fly data processing Ideal for closed loop feedback systems, such as adaptive optics See page 56 for Camera Link Output for Direct Data Access technical note EMCCD vs. CCD EMCCD CCD ü Single photon sensitive û 3 to 6 e - read noise û Multiplication noise ü No Multiplication noise ü Faster frame rates possible û Restricted to slower frame rates The basic trade-offs between EMCCD and conventional CCD amplifiers Ultra 888 vs. Ultra 897 Model CCD mode Pixel Readout Speeds Read Noise* ixon Ultra MHz, 1 MHz and 80 khz 3 e - ixon Ultra MHz, 1 MHz and 80 khz 3 e - * Read noise at slowest available readout speed for model The large field of view and the excellent resolution of the ixon Ultra 888 is ideal for the super-resolution images acquired with our home-built dstorm microscope. Dr. Alan Lowe and Dr. Ricardo Henriques, University College London, UK dstorm image of U2OS cells showing microtubules stained with Alexa 647 antibody captured with the ixon Ultra 888 model. Image courtesy of Dr. Alan Lowe and Dr. Ricardo Henriques, University College London, UK. Signal to Noise plots, comparing EMCCD mode at any readout speed vs CCD mode at 1 MHz and 80 khz readout speeds (the latter exclusive to ixon Ultra 897). Higher signal to noise can be secured with the 80 khz CCD amplifier for light levels greater than 12 photons/ pixel, but note that this corresponds to a max frame rate of 0.3 fps

14 ixon Software Solutions Third party software compatibility Andor Solis Solis is a ready to run Windows package with rich functionality for data acquisition and image analysis/ processing. Available on 32-bit and 64-bit versions of Windows (XP, Vista, 7 and 8). Andor Basic provides macro language control of data acquisition, processing, display and export. Andor SDK A software development kit that allows you to control the Andor range of cameras from your own application. Available as 32 and 64-bit libraries for Windows (XP, Vista, 7 and 8) and Linux. Compatible with C/C++, C#, Delphi, VB6, VB.NET, LabView and Matlab. Andor iq A comprehensive multi-dimensional imaging software package. Offers tight synchronization of EMCCD with a comprehensive range of microscopy hardware, along with comprehensive rendering and analysis functionality. Modular architecture for best price/ performance package on the market. Bitplane Imaris Imaris delivers all the necessary functionality for visualization, segmentation and interpretation of multidimensional datasets. By combining speed, precision and intuitive ease-ofuse, Imaris provides a complete set of features for handling multi-channel image sets of any size up to 50 gigabytes. Drivers are available so that the ixon range can be operated through a large variety of third party imaging packages, including: Metamorph (Molecular Devices Corporation) NIS Elements (Nikon) LAS (Leica) Xcellence (Olympus) Image Pro (Media Cybernetics) MicroManager (University of California, SF) Till Photonics Live Acquisition (Till Photonics) Imaging Workbench (Indec) WinFluor (University of Strathclyde) Maxim DL (Diffraction Limited) LabView (National Instruments) Matlab (MathWorks) Third Party Support For more information on third party software support and compatibility, visit Andor s web site and click on Software -> Third Party Software Matrix 3D images rendered by Imaris 26 27

15 The Andor Imaging Range Have you found what you are looking for? As an alternative to the ixon series, Andor offers an extensive portfolio of high performance low light imaging camera technologies. ikon CCD Clara Interline CCD Neo scmos ixon EMCCD Deep cooled, low noise CCD High-performance interline CCD Vacuum cooled, lowest noise scmos High performance EMCCD platform -100ºC cooling Industry lowest interline read noise (2.4 e - ) 1 electron read 30 fps Single photon sensitive and back-illuminated Back-illuminated > 90% QE -55ºC fan cooled; -40ºC vibration free mode 5.5 Megapixel / 6.5 μm Industry fastest frame rates 1 Megapixel to 4 Megapixel 1.4 Megapixel -40ºC vacuum cooling -100ºC cooling Enhanced NIR versions USB 2.0 true plug and play 30 fps sustained; 100 fps burst Flexible yet intuitive PV Inspector model (Optimized for EL / PL in-line inspection) USB 2.0 true plug and play Zyla scmos Fast, sensitive, compact, light scmos 4 GB on head memory 16-bit data range Fan off vibration free mode Quantify in electrons or photons 1 electron read 30 fps 5.5 and 4.2 Megapixel sensors / 6.5 μm 0ºC cooling at +35 C ambient 100 fps sustained (10-tap Camera Link) Cost effective USB 3.0 option 16-bit data range 28 29

16 Technical Notes Andor s high-performance EMCCD cameras have been the technology favorite of the vast majority of EMCCD-enabled laboratories across the globe. The following section is dedicated to providing a greater depth of understanding of the performance innovations underlying the ixon family of high-end EMCCD cameras, outlining the core technical reasons why Andor is still very much considered the EMCCD industry leaders, notably so in the key areas of sensitivity, speed, stability, longevity, quality and accessibility. Maximizing frame rate performance in EMCCDs UltraVac permanent vacuum head and performance longevity Count Convert - Quantifying data in Electrons and Photons New EMCCD Sensor Enhancements: Extended QE and Fringe Suppression Deep Vacuum TE Cooling and Darkcurrent Elimination OptAcquire Flexibility need not be complicated Minimizing Clock Induced Charge - finesse charge clocking Quantitative Stability in EMCCDs RealGain TM, Anti-Ageing and EMCAL photon counting in EMCCDs Fast Kinetics Mode Dynamic Range and EMCCDs Uncovering the Facts Making Sense of Sensitivity ixon Ultra and ixon3 Trigger Modes ixon Ultra Camera Link Output for Direct Data Access Because the vesicles are very small, the light sensitivity of the camera has to be as high as possible. This camera is the most sensitive available in our experience. It s better than any competitor around. Dr Roberto Zoncu, School of Medicine, Yale University, USA Commenting on the use of the ixon3 897 model in imaging fluorescently labeled endosomes in living cells using TIRF illumination

17 Technical Note Maximizing Frame Rate Performance in EMCCDs The ixon Ultra is capable of market leading frame rate performance, achieved from overclocking the horizontal and vertical shifts during readout while maintaining quantitative stability. Furthermore, the fastest possible continuous Region of Interest (ROI) frame rates can be attained using Crop Mode. Part 1. Overclocking horizontal and vertical shifts for fastest speeds Maximum frame rate performance in EMCCDs is a function of two parameters: (1) Pixel Readout Speed (horizontal); (2) Vertical Clock Speed. The former dictates how rapidly charge is pushed horizontally through the EM gain register and the remaining readout electronics, while the latter dictates the speed at which charge is vertically shifted down through both the exposed sensor area and masked frame transfer area of the chip. Significant advantages are gained through optimizing the camera electronics to enable both horizontal and vertical shifts to be speeded up through overclocking. ixon offers industry fastest vertical shift speeds, resulting in faster frame rates and reduced smearing, and is significantly faster under commonly employed conditions of sub-array/binning. Notably, the ixon Ultra markedly overclocks the pixel readout speed to up to as fast as 30 MHz, compared to the standard 10 MHz speed, further boosting the frame rate by 3x, in the case of the ixon Ultra 888. ixon Ultra 888 overclocked to 30 MHz pixel readout speed: 3x faster full frame rate Fastest vertical shift speeds yield further speed gains with ROI / binning Minimized smearing through faster vertical shifts Part 2. Boosting ROI Frame Rates with Crop Mode The ixon family offers Crop Mode (Figure 1), which carries the following advantages: Specialized readout mode for achieving very fast ROI frame rates (sub-millisecond exposures) from standard cameras. Continuous rapid spooling of images/spectra to hard disk. User selectable cropped sensor size highly intuitive software definition. The ixon Ultra is equipped with Optically Centred Crop Mode ROIs can be selected from the centre of the image. Ideal for super-resolution microscopy, ion signaling, voltage sensitive dyes and adaptive optics. The ixon is now available with the complementary OptoMask accessory, which can be used to shield the region of the sensor outside of the cropped area. If an experiment demands fast temporal resolution, but cannot be constrained by the maximum storage size of the sensor (as is the case for Fast Kinetics Mode of readout), then it is possible to readout the ixon in Crop Mode. In this mode, the user defines a ROI region, from the output corner of the sensor area (optically centred possible with ixon Ultra), ensuring to position the sample such that the ROI encompasses the region of the image where change is rapidly occurring (e.g. a calcium spark within a cell). The sensor subsequently imagines that it is of this smaller defined array size, achieved through the camera executing special readout patterns, and reads out at a proportionally faster frame rate. The smaller the defined array size, the faster the frame rate achievable. The reason that an ROI in Crop Mode can be significantly faster than a standard ROI is that Crop Mode overcomes the additional readout overheads associated with the reading and dumping signal charge from the unwanted pixels of the sensor area. In order to use Crop Mode, one has to ensure that no light is falling on the light sensitive area outside of the defined region. Any light collected outside the cropped area could corrupt the images that were acquired in this mode. For microscopy set-ups, this is now aided with an accessory called OptoMask, available from Andor. Table 1 shows frame rates that are achievable with the ixon Ultra 897 when in Crop Mode (corner tethered). Optically Centred Crop Mode: Enabling live cell super-resolution The ixon Ultra now comes with Optically Centred Crop Mode, which gives the user the option to break away from the corner tethered requirement of standard crop mode and select a number of pre-defined ROIs that are located in the centre of the image field. This is achieved with only minimal sacrifice in achievable frame rate; for example, in the case of the ixon Ultra 897, a 128 x 128 optically centred ROI delivering 569 fps. Optically centring of the ROI makes this mode extremely appealing to a number of microscopy techniques, including pointillism live cell super-resolution microscopy. For example, the ixon Ultra 897 can be operated in full 512 x 512 resolution at a frame rate suited to generation of fixed cell super-resolved images, then Optically Centred Crop Mode can be invoked with a 128 x 128 ROI for generation of super-resolved live cell images showing dynamic events. Table 2 shows frame rates possible from Optically Centred Crop Mode for a range of ROI sizes. Crop Mode is ideally suited to a number of challenging applications across many diverse fields of research. In terms of technology match, Crop Mode is particularly well suited to the many dynamic applications of EMCCD cameras. The fundamental advantage and a distinct feature of EMCCD technology is its ability to virtually eliminate the camera readout noise detection limit at any readout speed. This allows EMCCD detectors to be successfully used for applications where raw sensitivity and exposure time requirements ultimately prevent the use of conventional CCD systems. In biological imaging, Crop Mode can be successfully used to enhance performance and throughput in super-resolution pointillism applications including STORM, PALM and PALMIRA. Imaging frame rates exceeding 500 fps can be readily achieved from a 128 x 128 ROI (optically centred in the case of the ixon Ultra), which enables the camera for temporal and spatial requirements required for live cell super-resolution measurements.. Crop Mode can also be employed to achieve extremely fast temporal resolution in ion signaling measurements, such as observing calcium sparks. Samples labeled with voltage sensitive dyes also benefit from extremely fast imaging, with thousands of frames per second not being uncommon. There is also potential to use cropped EMCCDs for Fluorescence Correlation Microscopy (FCS), using multiple points or light sheet illumination. The mode is furthermore suited to multi-spectral fluorescence confocal scanning, as an alternative to the arrays of PMTs that have traditionally been used in this approach. The greater than 90% Quantum Efficiency of the back-illuminated sensor, single photon sensitivity, array architecture and rapid pixel readout speed can be exploited to markedly improve this approach. The laser dwell-time should be set to coincide with the time to expose and read-out a short row of approximately 32 pixels - sufficient spectral channels to yield effective un-mixing of several known emitting dyes, resulting in a data cube of 512 x 512 x 32 (spectral) taking less than 1 second to generate. There is a clear sensitivity advantage of EMCCD pixels over the usually employed PMT-technology, which is circa five-fold in the blue-green and up to ten-fold in the red. There are many physical science applications that can benefit from the speeds of Crop Mode, such as Bose Einstein Condensation (BEC) or lucky astronomy. EMCCD-based adaptive optics, for which smaller format EMCCD sensors are often used, can also benefit from this mode of readout. The ixon Ultra can be operated with a 128 x 128 ROI at a rate of almost 600 fps, readily pushing to > 1000 fps with binning, frame rates that are adaptable even to planetary adaptive optics. Standard Region of Interest (ROI) Binning 256 x x x x x x 1 1 x ,857 2 x ,136-4 x , ,613 - Crop Mode Region of Interest (ROI) Binning 256 x x x x x x 1 1 x , ,074 2 x ,085 2, ,589-4 x ,802 3,577 1,050 2,682 - Table 1 Frame rates achievable by the ixon Ultra 897 under both Standard ROI and Crop Mode (corner tethered) (a) (b) Centralized ROI Size (no binning) Figure 1. Crop Mode: The active imaging area of the sensor is defined in a way that only a small section of the entire chip is used for imaging. The remaining area has to be optically masked to prevent light leakage and charge spill-over that would compromise the signal from the imaging area. By cropping the sensor one achieves faster frame rates because the temporal resolution will be dictated by the time it requires to read out small section of the sensor. The ROI can be defined either at the output corner (a), or selected from a pre-defined list of optically centred ROIs (ixon Ultra only). Frame Rate 256 x x x x x 64 1, x 32 3,024 Table 2 Frame rates achievable by the ixon Ultra 897 operated in Optically Centred Crop Mode

18 Technical Note UltraVac Permanent Vacuum Head and Performance Longevity Andor s UltraVac vacuum process was designed not only to facilitate deep TE cooling, but also to provide absolute protection of the exposed sensor. Unless protected, cooled sensors will condense moisture, hydrocarbons and other gas contaminants. Such contaminants are particularly damaging towards the detecting surface of backilluminated sensors. Exposed to such outgassed contaminants, the Quantum Efficiency of a back-illuminated EMCCD will decline proportionally. Furthermore, the sensor can fail if excessive condensation forms. It was these compelling reasons that drove Andor to develop permanent vacuum technology more than 15 years ago. Andor have indeed perfected a proprietary Permanent Vacuum Head (Figure 1), essential not only to optimize cooling performance, but also to ensure that the sensor is protected and that this performance is retained year after year. Only Andor have shipped thousands of vacuum systems, enabling us to unequivocally substantiate our longevity claims with real reliability data. A back-illuminated EMCCD sensor must be housed in a hermetically sealed vacuum head with minimized outgassing, otherwise both cooling performance and the sensor QE itself will degrade. Vacuum Window (Anti-reflection Coated) Permanent, All-Metal Seal Heat Removal Benefits of Permanent Vacuum Head: Sustained vacuum performance over many years operation proprietary process to minimize outgassing. Peak QE and cooling will not degrade. Benefit from a thoroughly proven solution. More than 15 years of shipping vacuum systems to the field and a negligible failure rate - MTBF (mean time between failure) figure of more than 100 years. No one else can make or substantiate this claim with real data. Performance improves because the temperature of the chip can be reduced significantly. Better cooling (down to 100ºC with an enhanced thermoelectric peltier design) translates into substantially lower dark current and fewer blemishes. Elimination of condensation and outgassing means that the system can use only a single entrance window, with double antireflection coating you can believe the QE curve. Prevent convection heat transport from the front window, which would otherwise lead to condensation on the outside window. Technical Note Count Convert - Quantifying Data in Electrons and Photons One of the distinctive features of the ixon family is the capability to quantitatively capture and present data in units of electrons or photons, the conversion is applied either in real time or as a post-conversion step. Photons that are incident on pixels of an array detector are captured and converted to electrons. During a given exposure time, the signal in electrons that is collected in each pixel is proportional to the signal intensity. In EMCCDs, the signal in electrons is further multiplied in the EM gain register. The average multiplication factor is selected in software from the RealGain TM scale. It can be desirable to directly quantify signal intensity either in terms of electrons per pixel or in terms of incident photons per pixel. However, during the readout process, array detectors must first convert the signal in electrons (the multiplied signal in the case of EMCCDs) into a voltage, which is then digitized by an Analogue to Digital Converter (ADC). Each Analogue to Digital Unit (ADU) is presented as a count in the signal intensity scale, each count corresponding to an exact number of electrons. Furthermore, the signal value in counts will sit on top of an electronic bias offset value. In the ixon this baseline is clamped at 100 counts. Therefore, in order to back calculate to the original signal in electrons, the electron to ADU conversion factor must be very accurately stored by the camera (which varies depending on the pre-amplifier gain selection chosen through software). Calculation of the signal as absolute electrons also requires knowledge of the bias offset and the EM gain. The calculation path is shown in Figure 1. Furthermore, knowledge of the Quantum Efficiency (QE) at each wavelength and light throughput properties of the camera window enables this process to be taken a step further, allowing the signal to be estimated in photons incident at each pixel. For this step, the user must input the signal wavelength. In fluorescence microscopy for example, this would correspond to the central wavelength defined by a narrow band emission filter matched to the fluorophore of interest. If the spectral coverage of the signal on the detector is too broad, such that the QE curve varies significantly throughout this range, then the accuracy of the incident photon estimation would be compromised. Counts Signal counts Signal electrons Original signal in electrons Original signal in incident photons Figure 1 - Count Convert calculation path Benefits of Count Convert (minus bias offset) (x sensitivity e - / count) ( EM gain) ( wavelength of interest) CCD TE Cooler Schematic of a Permanent Vacuum Head The Count Convert functionality of the ixon provides the flexibility to acquire data in either electrons or incident photons, with negligible slow down in display rate. Furthermore, the option exists to record the original data in counts and perform this important conversion to either electrons or photons as a post-conversion step, while retaining the original data. Quantify data in electrons or incident photons Convenient estimate of sample signal intensity at the detector Real time or post-convert Reference between different samples, users and set-ups Meaningful signal relating to PALM/STORM localization accuracy Figure

19 Technical Note New EMCCD Sensor Enhancements: Extended QE and Fringe Suppression Selected ixon models are now available with a new EX2 Dual Anti-Reflection coating and Fringe Suppression technologies, applied to the back-illuminated sensors. EX2 affords a significant enhancement of the Quantum Efficiency performance in the blue and red/nir regions, and Fringe Suppression reduces the effect of etaloning in the NIR. Quantum efficiency (%) EX2 BV +20ºC Quantum efficiency (%) EX2 BV -100ºC EX2 Technology Extended QE with Dual AR coating Available on the new speed-boosted ixon Ultra 897 and 888 cameras, EX2 technology facilitates broadening of the QE range of the backilluminated sensors through implementation of a new dual layer anti-reflection coating process, developed by sensor manufacturer e2v (Chelmsford, England). The net effect is to offer significantly improved sensitivity in both the blue and NIR wavelength regions, whilst maintaining ~ 90% QE across the remainder of the visible region. QE varies somewhat as a function of sensor cooling temperature, so it can be useful to view reference QE curves at two extreme temperatures to achieve a sense of this dependence. Figure 1 shows QE curves for the new EX2 dual AR coated sensors versus that of the standard BV single AR coating (mid-band optimized) at both +20 C and -100 C sensor cooling temperatures. It is clear, notably under more typical conditions of deep cooling, that the EX2 curve exhibits a marked widening of the QE response into both the blue and red/nir wavelength ranges, while maintaining ~ 90% QE in the important region in between. Figure 2 shows the differential variation in QE for the EX2 sensor relative to the standard BV curve, presented again for +20 C and -100 C sensor cooling temperatures. It is evident that a relative QE boost of up to 25% is possible in the blue region at -100 C. restrict the ability to perform high integrity quantitative imaging. Fringe Suppression refers to a mature technology offered by the sensor manufacturer e2v, and refers to a process whereby structured roughening is introduced on the sensor s back surface, which, when coupled with application of a given anti-reflection coating, leads to the reduction of the radiation reflections within the sensor, hence reducing significantly the fringing patterns. This does not affect other sensor characteristics such as dark noise, readout noise or readout speeds. The process has been validated over many years of CCD manufacturing and is now available on EMCCD sensors, specifically relating to the following ixon models: ixon Ultra 897 and ixon Ultra Wavelength (nm) Wavelength (nm) Figure 1: Back-illuminated EMCCD sensor QE curves, comparing standard BV mid-band AR coating versus new EX2 dual AR coating. Reference data available at +20 C and -100 C sensor cooling temperatures. 30% 25% 20% 15% +25C -100C Furthermore, when compared to a competitive multi-ar coated backilluminated EMCCD sensor (available on the same sensor formats), the EX2 sensor exhibits higher QE in the wavelength range between nm, i.e. exhibiting less QE drop off in this important visible region, and near identical NIR sensitivity. In terms of credibility, it should also be noted that the QE curves presented by Andor for EX2 and BV coatings are each measured by the sensor manufacturer e2v on the same apparatus, and thus are truly comparable. QE curves measured on different set-ups are subject to significant variation. QE 10% 5% 0% % Fringe Suppression Technology Fringing is generated in back illuminated sensors where the extreme flat surfaces of the substrate effectively create a Fabry-Perot etalon. The finesse (thickness) of the cavity is such that interference between reflected radiation within the sensor and/or the incoming radiation at the same wavelength can occur. The interference pattern level will be wavelength and cavity finesse dependant. This interference pattern can affect the acquired image or spectrum. Indeed, for NIR applications, such as imaging of Bose Einstein Condenates, etaloning effects, often observed more notably beyond ~ 750nm, can sometimes -10% Wavelength (nm) Figure 2: Differential variation in QE for the EX2 sensor relative to the standard BV curve for +25 C and -100 C sensor cooling temperatures

20 Technical Note Deep Vacuum TE Cooling and Darkcurrent Elimination -30ºC (A) -80ºC -30ºC (B) On harnessing EMCCD technology, dark current is an absolutely critical parameter to minimize, more so than in a standard sensitive CCD. The reason for this is that thermally generated electrons are amplified by EMCCD just as photon-generated electrons (signal) are amplified. For optimal sensitivity in EMCCDs, thermoelectric cooling of the sensor must be deep enough that this noise source is virtually eliminated. It is important to recognize that this point applies very much to short exposure operation also; the sensor readout process alone results in significant dark current production if left untreated. The above point is demonstrated by the following simple test: Figure 1(A) shows two dark images taken with a 512 x 512 back-illuminated sensor (from e2v) at two different cooling temperatures: -80ºC and -30ºC. All are in frame transfer mode at the maximum frame rate of 34 full frames per second, i.e. ~ 29 ms exposure. EM gain is set at x1000; at such an EM gain setting the vast majority of dark current and CIC events will be exposed. on a light tight imaging chamber using weak LED illumination through pinholes. The light levels used are typical of an experiment involving imaging of weak luminescence signal. The photon flux is so low, that a two minute exposure is required in order to visualize the pinhole signals. It is clear from the significantly improved SNR, and therefore contrast, at -95ºC cooling, that such extremely low cooling temperatures are recommended for longer exposure acquisitions. 1 MHz EM mode readout limits background of long exposures When back-illuminated EMCCD sensors from e2v are read out at faster horizontal speeds, such as 10 MHz, the EM-amplified background becomes notably higher over longer exposures. A slower 1 MHz readout mode is required to minimize background under these conditions. Figure 1: (A) shows dark images taken at x1000 gain at different cooling temperatures, 29 ms exposure time. Vertical shift speed was 0.5 μs/row to ensure minimal CIC. (B) shows typical line intensity profiles across a row of 512 pixels, taken from such dark images at three different cooling temperatures. The cleanest noise floor is clearly seen under conditions of deep cooling, even for such short exposure times. -70ºC -95ºC -50ºC -85ºC The speckled salt and pepper noise pattern that is quite obvious in the -30ºC condition is due almost entirely to amplified dark current electrons. Note that each of these images have ALREADY been optimized for minimal CIC, so only the effect of cooling is being demonstrated. If CIC had not been minimized, the 30ºC situation would have appeared bleaker still, with an extremely dense EMamplified noise floor. As an alternative view Figure 1(B) shows a line intensity plot from a single row from each image. The readout noise is visible as the fuzzy baseline of each trace with the dark current spikes sticking out of it; this is just what we expect. These spikes, or background events, are what set the remaining detection limit of the camera, not the readout noise. It is instantly clear that cooling is beneficial. The performance at 80ºC is by far the best and there is no way that this level of low noise detection limit can be achieved at 30ºC, under any circumstances. It is important to note that even with a short exposure time, the -30ºC background events are predominantly from dark current. As such, we are clearly better off with much deeper cooling, regardless of exposure time. Single thermal electrons are amplified by the EMCCD gain mechanism. Deep vacuum TE cooling is critical to optimize the sensitivity performance of backilluminated EMCCD sensors, otherwise the raw sensitivity will be compromised, even under conditions of short exposures. This is not so apparent for shorter exposure times (e.g. < 1 sec) but when you go towards multiple secs to mins it is certainly apparent. Some example dark noise images are shown in Fig 1 and Fig 2 for two different cooling temperatures, taken with the ixon 888 camera at x1000 EM gain. This basic trend can indeed be considered an oddity, because you would intuitively imagine that the readout speed should not influence the noise background that is built up during the exposure, but nevertheless it is fact. This means that we recommend using a readout speed of 1 MHz for longer exposure times. This is not really disadvantageous since you have already compromised the frame rate with the longer exposure time. For shorter exposure times/faster frame rate measurements, it is still best to opt for 10 MHz readout speed because dark current (being time dependent) is lower and CIC is the prominent contributor. Since CIC is elevated by use of slower readout parameters, then you are better using 10 MHz. Also, you tend to need the faster readout speeds to achieve faster frame rates. Recommended Pixel Readout Speed Cooling Temperature -70ºC -100ºC 17 MHz < 1 sec < 10 sec 10 MHz < 1 sec < 10 sec 1 MHz > 1 sec > 10 sec 60 sec 1 MHz 60 sec 10 MHz Figure 2: Images of extremely weak LED signal (signal intensity typical of weak luminescence experiments) acquired with ixon 888 at cooling temperatures -70ºC and -95ºC (water cooling to achieve latter), 120 sec exposure times, subregion show. The need to push to such deep cooling temperatures can be readily observed under such extreme low light conditions. Deep TE cooling can also make a tangible difference to signal to noise ratio for longer exposure conditions. Figure 2 shows extremely low light images, recorded at -70ºC and -95ºC with the ixon 888 A slow readout 1 MHz mode is required to minimize EM-amplified background under longer exposure acquisition conditions. Figure 3: Dark noise background at 1 MHz and 10 MHz readout speeds; 60 sec exposure and -70ºC cooling

21 Technical Note OptAcquire Flexibility need not be complicated OptAcquire is a unique control interface, whereby a user can conveniently choose from a pre-determined list of set-up configurations. Each is designed to optimize the camera for different experimental acquisition types, thus removing complexity from the extremely adaptable control architecture of the ixon. The control architecture of the ixon is extremely tunable, meaning the camera can be adapted and optimized for a wide variety of quantitative experimental requirements. They range from fast single photon counting through to slower scan, 16-bit Dynamic Range measurements. However, successfully optimizing EMCCD technology is not a trivial exercise, with various set-up parameters directly influencing different camera performance characteristics. OptAcquire has been designed as a unique interface whereby a user can choose from a pre-determined list of nine camera set-up configurations. A variety of set-up parameters are balanced behind the scenes through the OptAcquire menu. Furthermore, advanced users may wish to create their own additional OptAcquire modes to aid future set-up convenience. ixon control parameters include: EM gain this parameter has a direct bearing on both sensitivity and Dynamic Range. Vertical clock speed flexibility in this parameter is critical to optimizing the camera for lowest noise, fastest speed, minimal frame transfer smear or maximum pixel well depth. Pre-defined OptAcquire modes: Sensitivity and Speed (EM Amplifier) Dynamic Range and Speed (EM Amplifier) Vertical Clock Amplitude can be employed to help over-clock the sensor to achieve faster frame rates. It can also be used to reduce charge leakage into the image area when there is saturated signal in the frame transfer storage area (e.g. when combining very short exposure with a slow readout speed). Horizontal readout speed ranging between faster frame rates and best Dynamic Range. Pre-amplifier gain trading off reduced digitization noise versus accessing full pixel well depth. EM / Conventional amplifier to choose between ultra-sensitive EMCCD operation or traditional high Dynamic Range CCD operation. The latter is recommended for relatively brighter signals or when it is possible to apply long exposures to overcome read noise floor. Frame Transfer (overlap) overlap readout is used to achieve 100% duty cycle, ideal for fastest frame rate measurements without switching exposure time between frames. This mode should be deselected for time-lapse experiments. Optimized for capturing weak signal at fast frame rates, with single photon sensitivity. Suited to the majority of EMCCD applications. Configured to deliver optimal Dynamic Range at fast frame rates. Moderate EM gain applied. Technical Note Minimizing Clock Induced Charge - Finesse Charge Clocking Clock Induced Charge (CIC) can be considered the remaining detection limit in EMCCD and must be minimized. Careful and rapid clocking are crucial contributors to achieving this. The remaining limiting factor for EMCCD sensitivity, provided dark current has already been minimized through effective TE cooling, is a spurious noise source called Clock Induced Charge (CIC). This form of electron generation can occur even under normal clocking in any CCD, but when properly optimized the rate of occurrence is very small; i.e. CIC occurrence can be minimized down to the order of 1 in 200 pixels. For an EMCCD at high EM gain, such individual electrons can be seen as sharp spikes in the image and any CIC will become visible. For several years Andor has had very fine nanosecond resolution over EMCCD clockings and is well aware of the important parameters for reducing CIC, such as fine temporal control over the clock edges. Furthermore, there is a well-established direct link between pushing vertical clocks faster and achieving lower CIC. This clock speed dependence is amply illustrated in the following basic test. Figure 1 shows a series of dark images and corresponding intensity profiles (across a random row from each image), using 30 ms exposures and EM gain of x1000. The cooling temperature in each case was set at -85ºC to ensure virtual elimination of dark current contribution, so any amplified noise spikes are derived primarily from the remaining CIC. 3.3 µs 1.7 µs You can see that the vertical clock speeds have a significant impact on further reducing CIC, with the speed of 0.5 μs/shift offering the lowest CIC and therefore highest EMCCD sensitivity. This ixon3 clock speed performance is faster than other EMCCDs in the industry. After having minimized dark current through < -80ºC cooling, the remaining detection limit in back-illuminated EMCCDs is given by the number of Clock-Induced Charge noise events. Andor s industry-exclusive combination of high resolution clocking parameters and sub-microsecond clock speeds are fundamental to minimizing CIC, enabling truly high-end EMCCD sensitivity to be claimed. Clock Induced Charge Comparison Fastest Frame Rate (EM Amplifier) For when it s all about speed. Optimized for absolute fastest frame rates of the camera. Especially effective when combined with sub-array/binning selections. Time Lapse (EM Amplifier) Configured to capture low light images with time intervals between exposures. Overlap ('frame transfer') readout is deactivated. Time Lapse and Short Exposures (EM Amplifier) Configured to minimize vertical smear when using exposure times < 3 ms. 0.5 µs EMCCD Highest Dynamic Range (EM Amplifier) Combines EMCCD low light detection with the absolute highest Dynamic Range that the camera can deliver. Since this requires slower readout, frame rate is sacrificed. (B) CCD Highest Dynamic Range (Conventional Amplifier) photon counting photon counting with Long Exposures (> 10 sec) Optimized for slow scan CCD detection with highest available Dynamic Range. Recommended for brighter signals or when it is possible to apply long exposures to overcome noise floor. Configuration recommended for photon counting with individual exposures < 10 sec. Configuration recommended for photon counting with individual exposures > 10 sec. (A) Figure 1 (A) shows dark images taken at x1000 gain at different vertical shift speeds, 29 ms exposure time. Cooling temperature was -85ºC to ensure minimal dark current contribution. (B) shows typical line intensity profiles across a row of 512 pixels, taken from such dark images at three different vertical shift speeds. The cleanest noise floor is clearly seen under conditions of faster vertical shifts, an exclusive Andor capability

22 Technical Note Quantitative Stability in EMCCDs EMCCDs are susceptible to various sources of data instability. Each of these sources have been addressed in ixon Ultra and ixon3 to ensure reliable quantitative performance throughout a kinetic acquisition and also repeatability between measurements. Baseline Clamp The baseline (or bias level) is an electronic offset added to the output signal from the EMCCD sensor to ensure that the displayed signal level is always a positive number of counts. No actual noise is associated with this positive counts value and thus it is important to recognise that it does not affect sensitivity. However, one must remember to subtract the baseline offset value from the signal intensity when performing signal to noise calculations. Traditionally, when acquiring data, small changes in heat generation of the driving electronics within the detector head may cause some drift in the baseline level. This is often particularly observable during long kinetic series. Since 2002, Andor has offered a solution to this undesirable effect in our high-end EMCCD cameras. Any drift in the baseline level can be corrected by using our innovative Baseline Clamp option. Baseline Clamp corrects each individual image for any baseline drift by subtracting an average bias signal from each image pixel and then adding a fixed value to ensure that the displayed signal level is always a positive number of counts. As such, the baseline remains at a rock-steady value during a fast kinetic series, as shown in Figure 1. Note: The baseline bias level is also susceptible to variation at different EM gain settings. Again the ixon baseline clamp corrects for this, ensuring the bias level is clamped no matter what EM gain setting is selected. Electron Multiplication (EM) and Temperature Stability It is a well recognized fact that EM multiplication factor is temperature dependent. That is why ixon pioneering RealGain TM - a linear and quantitative EM gain calibration - is temperature compensated, i.e. the same precise correlation between software EM gain selection and actual EM gain holds at whatever cooling temperature is selected. This will be addressed in more detail later. Kinetic series were recorded over fps (17.8 ms exposure time), in frame transfer mode, with Baseline Clamp activated (such that absolute bias stability is in place also). A moderately intense signal, such that instability would not be lost in the signal shot noise, was recorded with x300 EM Gain. A Region of Interest kinetic plot was derived from the data, as shown in Figure 2. Significantly, there is no additional relative signal variation observable over the duration of the kinetic series. Be careful of light source instability When performing stability measurements, care must be taken to assess the stability of the light source. As an example, an ixon 897 was mounted on a research grade widefield epifluorescence microscope and fixed cells, immuno-stained with fluorescence dyes (Invitrogen Molecular Probes) were imaged over a similar kinetic series, as shown in Figure 3. In these experiments the variation of the signal intensity was significant, whether EM gain was on or off; the amplitude of variation was much higher than the shot noise on the signal. This indicates that the light source of the microscope itself can often be subject to much greater stability variations than could be derived from any EM gain instability observations. One needs to very carefully check for all sources of undesirable signal fluctuation, such as stability of illumination (background illumination fluctuations can also contribute) when conducting quantitative time course experiments. Baseline Clamp ON Mean ROI signal / ADU Figure 1 - ixon Baseline Clamp (bias stability) in operation. A standard feature in ixons since Average ROI signal / counts Baseline Clamp OFF Mean ROI signal / ADU ixon Ultra fps; EM gain x 300 Frame Number Figure 2 EM gain stability in the back-illuminated ixon Ultra 897 (512 x 512 pixels) LED measurement. However, another by-product of this temperature dependence is that we must also pay close attention to optimized temperature stability regulation. The ixon Ultra maintains a thermostatic precision of +/ C. The results of such attention to detail are best understood through observation of the stability via measurement of a stable light source, as shown in Figure 2. Here we used the back-illuminated ixon Ultra 897 to measure signal from a stable LED source overlayed with a resolution chart, imaged in conditions of zero ambient background light (use of a light tight darkbox). Average ROI signal / counts EM Gain ON Time / sec Average ROI signal / counts EM Gain OFF Time / sec Figure 3 Instability of the light source on a standard research grade epifluorescence microscope

23 Technical Note RealGain, Anti-Ageing, EMCAL and Longevity In 2006 Andor once again raised the bar by introducing some significant new technology innovations. These particular pioneering steps were to set new high standards in quantitative EMCCD usage and general EMCCD longevity expectations, which others in the industry are only now beginning to adopt. Part 1: RealGain, Anti-Ageing and EMCAL RealGain Linear - In response to considerable demand from our customers, Andor has set about a detailed analysis of the EM voltage dependence, and has successfully converted the non-linear relationship between EM gain and the EM software setting into a linear one. Real - Importantly, the true EM gain (i.e. the absolute signal multiplication factor) is selected directly from the software linear gain scale, as shown in Figure 1. No more guesswork with arbitrary gain units across a non-linear scale - the gain you ask for is the gain you get. Select the best gain to overcome noise and maximize Dynamic Range. Temperature Compensated Although EM gain is temperature dependent, Andor s real/linear gain calibration extends to any EMCCD cooling temperature. Selecting x300 gain software setting at -50 C, or at 100 C gives the same x300 true EM gain. Importantly, this means that there is no need to recalibrate EM gain on each use in multi-user laboratories and facilities. Anti-Ageing of EM gain It has been observed that EMCCD sensors, more notably in cameras that incorporate L3Vision sensors from e2v, are susceptible to EM gain fall-off over a period of time. This phenomenon has been documented by e2v and can be viewed on their web site. All backilluminated EMCCD sensors are of this brand and therefore all are susceptible to EM gain ageing. It is important to note that the ageing effect applies to any EMCCD camera, by any manufacturer, that incorporates these L3Vision sensors. In Andor s ixon range, this corresponds to the ixon3 860 model and the ixon Ultra 897 and 888 models. If left unchecked, this ageing phenomenon has the potential to significantly compromise the long-term quantitative reliability of EMCCD cameras. Fortunately, if these highly sensitive sensors are integrated into intelligently designed camera electronics, ageing can be minimized and should not present any real problem to the user. Andor has recognized the ageing issue and has been busy implementing innovative measures to stabilize the EM gain on these sensors, ensuring that this ground-breaking ultra-sensitive technology can deliver a prolonged quantitative service to the user. ixon cameras have been internally configured to ensure that the rate of EM gain fall off is significantly reduced under standard operation. Part of the measures taken has been to invoke real EM gain limits, coupled with signal intensity feedback warnings (after EM amplification) to ensure that the user is more restricted in his/her ability to apply excessive EM gain and/or signal. The EM gain scales offered are more than sufficient to render the read noise floor as negligible for a given signal intensity and readout speed. These tight user restrictions significantly reduce the rate of EM gain fall off. EMCAL Andor have developed a unique and patented method of user-initiated EM gain self-recalibration. Even after exercising due care during usage and availing of the above anti-ageing restrictions, the EM gain may deplete over an extended period of time. The EMCAL selfrecalibration process is very easily initiated by the user. At the touch of a button, a routine is triggered that measures EM gain and uses the ixon in-built temperature compensated RealGain TM scales to reset the EM gain calibration (if required), to deliver the true values requested on the software scale i.e. resetting the factory calibration. EMCAL stands to markedly prolong operational lifetime and quantitative reliability of the technology, and circumvents the need to return the camera to the factory for recalibration. Figure 1 RealGain TM calibration in the ixon the same linear relationship holds across all cooling temperatures. Part 2: Longevity How extensively can Andor s back-illuminated EMCCDs be used before they can no longer be recalibrated (EMCAL ) to factory EM gain settings? One common concern associated with the EM gain ageing phenomenon and the associated EMCAL recalibration fix is that of longevity. The clock voltage setting, which must be adjusted as part of the recalibration routine, will eventually reach a maximum threshold value, after which further rescaling is not possible and EM gain will then fall off irreparably upon further extensive use of the camera. The question is, when is this likely to happen under typical use? Andor ran extensive testing on the ixon 897 camera in order to project the operational lifetime of the gain register (where signal amplification occurs on-chip) in back-illuminated EMCCDs, the conditions of testing are described below: Overall duration of test: ~14 months Camera usage during this period: Continuous 24 hours per day, 7 days per week Frame rate: 30 frames / sec EM gain setting: x1000 Photons / pixel / frame: 90 Number of pixels illuminated: ~ 200,000 (~ 75% of array) EMCAL: Applied once per day. 7 Years Figure 2 Ageing profile of an Andor backlit EMCCD. Test conditions: 24/7 operation; 30 fps; x90,000 electrons per pixel through gain register; ~ 200,000 pixels illuminated Results The clock voltage required to maintain the EM gain calibration was measured once daily using Andor s EMCAL routine and the plot shown was generated by extrapolating from the equation derived from the gain ageing trend. This shows that under the test conditions employed, the EMCCD calibration would only be expected to reach ~ 65% of the available clock voltage scale after seven years of continuous operation. Demanding test conditions The combination of parameters employed in this test represents quite aggressive acquisition conditions. ~1.1 billion images were recorded during this period, with ~ 200,000 illuminated pixels per image, corresponding to ~ 220,000 billion pixels being amplified through the gain register with x1000 EM amplification per pixel. Most users would not subject the camera to 24/7 continuous acquisition at 30 fps. We also strongly recommend that, except for photon counting, the EM gain setting is limited to no more than x300 for the vast majority of applications, but x1000 was chosen here as a more rigorous test condition. Finally, it is quite rare that 75% of all pixels in the array will be subjected to uniform signal of this magnitude, as was imposed on the sensor here. In reality the light emitting species of typical user samples will project onto a much smaller fraction of pixels from frame to frame. EMCAL does NOT accelerate EM gain ageing It is important to recognize that the rate of ageing is not accelerated by routine application of Andor s EMCAL routine. The rate of ageing is determined by the illumination and EM gain conditions that the sensor is subject to through operation, irrespective of routine recalibration using EMCAL. If the previous mechanism of EM gain recalibration were to be used, whereby the camera is shipped back to the factory less frequently for manual readjustment of the clock voltage, the progress along the ageing curve would not be any different from that shown here (adjusting for the additional time that the camera would be out of action). Conclusion For the vast majority of low-light applications and taking due care and attention to stay within recommended operating conditions, applying EMCAL as required, the gain ageing phenomenon is not considered to ever impose a restriction on the quantitative reliability of your Andor ixon camera

24 Technical Note Photon Counting in EMCCDs Photon counting in EMCCDs is a way to overcome the multiplicative noise associated with the amplification process, thereby increasing the signal to noise ratio by a factor of root 2 (and doubling the effective Quantum Efficiency of the EMCCD). Only EMCCDs with a low noise floor can perform photon counting. The approach can be further enhanced through innovative ways to post process kinetic data. Figure 1 Photon Counting vs. Standard EM-on Imaging for very weak signals: Images A, B and C were recorded under identical illumination conditions, identical exposure times and each with EM gain set at x1000. The benefit of photon counting under conditions of low Clock Induced Charge (CIC) are evident. Images D and E are derived from a larger number of accumulated images, to yield a greater measurable signal to noise ratio. An identically positioned Region of Interest on each image was used to determine S/N of 3.86 and 6.02 for standard and photon counted images respectively. This factor improvement is in accord with the theory of photon counting circumventing the influence of multiplicative noise (noise factor) in EMCCD signals. The industry-leading dark current and Clock Induced Charge (CIC) specification of Andor s back-illuminated ixon Ultra 897 and 888 render them uniquely suited to imaging by photon counting. Photon counting can only be successfully carried out with very weak signals because, as the name suggests, it involves counting only single photons per pixel. If more than one photon falls on a pixel during the exposure, an EMCCD (or an ICCD for that matter) cannot distinguish the resulting signal spike from that of a single photon event, and thus the Dynamic Range of a single frame exposure is restricted to one photon. To successfully photon count with EMCCDs, there has to be a significantly higher probability of seeing a photon spike than seeing a dark current/cic noise spike. The ixon Ultra 897 and 888 have the lowest dark current/cic performance on the market, yielding both lower detection limits and higher contrast images. Under such ultra-low light conditions, photon counting mode imaging carries the key benefit that it is a means to circumvent the multiplicative noise, also know as Noise Factor. Multiplicative noise is a by-product of the Electron Multiplication process and affects both EMCCDs and ICCDs. In fact, it has been measured to be significantly higher in ICCDs. The noise factor of EMCCDs is well theorized and measured; to account for it you increase the shot noise of the signal by a factor of square root 2 (~x1.41). This gives the new effective shot noise that has been corrected for multiplicative noise. The effect of this additional noise source on the overall Signal to Noise ratio can be readily viewed in the S/N plots in the technical note entitled EMCCD signal to noise plots. Photon counting mode does not measure the exact intensity of a single photon spike, it merely registers its presence above a threshold value. It does this for a succession of exposures and combines the individual binary images to create the final image. As such, this mode of operation is not affected by the multiplication noise (which otherwise describes the distribution of multiplication values around the mean multiplication factor chosen). The end result is that low light images acquired through this mode of acquisition are improved by a factor of ~x1.41 Signal to Noise, compared to a single integrated image with the same overall exposure time. To successfully photon count with EMCCDs, there has to be a significantly higher probability of seeing a photon spike than seeing a dark current/cic noise spike. The lower the contribution of this spurious noise source to a single exposure within the accumulated series, the lower the detection limit of photon counting and the cleaner the overall image will be, as demonstrated in Figure 1. The ixon Ultra 897 and 888 have the most effective combined cooling/cic minimization on the market, lower than other competing EMCCDs utilizing the same 512 x 512 or 1024 x 1024 sensors. As such, the detection limit for photon counting is markedly lower. The ixon intuitively offers photon counting modes, either as a real time acquisition or as a post-processing step. OptAcquire can be used to first optimize the camera for photon counting acquisition. Photon Counting by Post-Processing As a post-processing analysis, the user holds the flexibility to trial and error photon count a pre-recorded kinetic series, trading-off temporal resolution vs SNR by choosing how many images should contribute to each photon counted accumulated image. For example, a series of 1000 images could be broken down into groups of 20 photon counted images, yielding 50 time points. If it transpires that better SNR is required, the original dataset could be re-treated using groups of 50 photon counted images, yielding 20 time points. TIME Figure 2 Schematic illustration of how photon counting can be applied to a kinetic series as a post processing step, affording increased flexibility in trial and error trading SNR vs temporal resolution

25 Technical Note Fast Kinetics Mode Fast Kinetics Acquisition Mode can be used to acquire bursts of data with submicrosecond time resolution. The ixon family is configured to make available not only the rows of the image area, but also rows under the frame transfer mask for storage of acquired data prior to readout. The overclocked vertical shift speeds of the ixon renders it ideal for extremely fast temporal resolution in Fast Kinetics Mode. Fast Kinetics is a special read out mode of ixon Ultra and ixon3 that uses the actual sensor as a temporary storage medium and allows an extremely fast sequence of images to be captured. The capture sequence (Figure 1) is illustrated here: Step 1: both the Image and Storage areas of the sensor are fully cleaned out (the Keep Clean Cycle) Step 2: the Keep Clean Cycle stops and the acquisition begins. The image builds up on the illuminated section of the sensor. Step 3: the sensor remains in this state until the exposure time has elapsed, at which point the complete sensor is clocked vertically by the number of rows specified by the user. Steps 4 and 5: the process is continued until the number of images stored equals the series length set by the user. The ixon is set up to utilize the entire area of the frame transfer mask for additional signal storage. Step 6: at this point the sequence moves into the readout phase by first vertically shifting the first image to the bottom row of the sensor. The sensor is then read out in the standard method. Points to consider for Fast Kinetics Mode: Light MUST only be allowed to fall on the specified sub-area. Light falling anywhere else will contaminate the data. The maximum number of images in the sequence is set by the position of the sub-area, the height of the sub-area and the number of rows in the CCD (Image and Storage area). There are no Keep Cleans during the acquisition sequence. The industry fastest vertical shift speeds of the ixon enables fastest time resolution with minimal vertical smearing. A range of internal trigger and external trigger options are available for Fast Kinetics Readout. The Fast Kinetics Mode capability of the ixon Ultra and ixon3 renders it uniquely suited to microsecondorder kinetic measurements, facilitated by rapid vertical shifts and extensive on-chip storage areas. Figure 1 Illustration showing Fast Kinetics Mode capture sequence of the ixon 48 49

26 Technical Note Dynamic Range and EMCCDs Uncovering the Facts Calculating Dynamic Range in EMCCDs has often been a source of confusion, due to the additional requirement to factor in EM gain and the extended well capacity of the gain register. High Dynamic Range can be accessed in EMCCDs with careful fine tuning of EM gain. Dynamic Range (DR) is given by: Full well capacity DR = Detection limit Calculating Dynamic Range in an EMCCD camera is a slightly more complicated story than for conventional CCDs. This is because of the favorable effect of EM gain on the detection limit vs. the limiting effect of EM gain on the full well capacity. The easiest way to address this is to first take each parameter separately: Detection Limit and EM gain The main function of EMCCD is to eliminate the read noise detection limit and enable detection of weak photon signals that would otherwise be lost within this noise floor. With EM gain, the detection limit is given by the Effective Read Noise, i.e. the read noise divided by the gain multiplication, down to one electron. Why never less than one? This stems from the definition of detection limit, which is essentially the signal equal to the lowest noise level. Since you can t get a signal less than one photon, then the detection limit should never be taken as less than one electron. For example, the ixon Ultra 888 has a read noise of ~45 MHz with EM gain off. At EM gain x2, the new detection limit can be considered to be 22.5 electrons effective read noise, at x5 it will be 9 electrons, at x45 it will be 1 electron. At x100, the Effective Read Noise will be 0.45 electrons, but as far as the Dynamic Range calculation is concerned, this detection limit must still be treated as 1 electron. Full Well Capacity and EM gain One might imagine that applying EM gain will decrease the full well pixel capacity proportionally. This is indeed the case, but a buffer has been built into EMCCD cameras to enable at least some EM gain to be applied while maintaining the original well capacity. This buffer is in the form of a higher capacity in the gain register pixels, where the multiplication actually takes place. So, the true capacity is given by the capacity of the pixels of the sensor, but as you apply EM gain this holds only up until the point where the larger capacity of the gain register pixels also become saturated by applied EM gain. After that point, you have to correct the effective full well of the sensor to be equal to the full well of the gain register divided by the gain. Dynamic Range and EM gain These above factors combined mean that as EM gain is increased, ixon Ultra 888 back-illuminated EMCCD can be read out at either 30, 20, 10 or 1 MHz speeds. This offers extended flexibility to balance Dynamic Range vs. frame rates. Furthermore OptAcquire can be used to select optimal Dynamic Range settings at fastest and slowest speeds. Dynamic Range will increase with gain to a maximum, level off and then reach a point at which it begins to deplete again with further gain. This can seem complicated, but fortunately these DR vs EM gain relationships can be readily plotted out and visualized in graphical form, as exemplified in Figure 1. There are a number of interesting points to note from these plots: 1. The rationale behind offering readout speeds slower than 17 MHz through the EM-amplifier is so that frame rate can be traded off against Dynamic Range. You can see that the highest Dynamic Range through an EM amplifier comes from the slowest 1 MHz readout speed. 2. At any readout speed through the EM-amplifier, the best combination of Dynamic Range and sensitivity can be obtained at an EM gain setting equal to the readout noise at that speed. At this point the DR is at maximum and the effective readout noise is 1 electron (i.e. just on the verge of single photon sensitivity). 3. At x1000 EM gain the Dynamic Range is only 400:1. Excessively high EM gain can also accelerate EM gain ageing in back-illuminated EMCCDs. EM gains of x300 or less are more than sufficient to optimize sensitivity, while ensuring Dynamic Range is not excessively compromised. The only occasions when Andor recommends extending EM gain to x1000 is for single photon counting experiments. 4. The highest Dynamic Range is through the conventional CCD amplifier. 5. It is clear that the actual sensor Dynamic Range only exceeds 14-bits at 1 MHz, through either EM or a conventional amplifier. Therefore, it is at 1 MHz that it becomes important to match this higher Dynamic Range output with a scientific grade, noise free 16- bit A/D digitization. Note: There is a direct relationship between readout noise and maximum Dynamic Range at a given readout speed. Lower readout noise affords higher Dynamic Range. The readout noise specification used in calculating Dynamic Range must be with EM gain turned off, as quoted in all ixon spec sheets. We note, however, that another prominent EMCCD provider chooses to quote their lowest read noise value, not for EM gain-off, but only for EM gain x4, x6 or x10 (model dependent). In this case, to arrive at the real read noise spec you would have to multiply the quoted figure by x4, x6 or x10. Figure 1 Dynamic Range vs EMCCD gain, for ixon Ultra 897. Shown for EM 17, 10 and 1 MHz readout speed and for Conventional amplifier at 1 MHz readout speed. Well capacities used in DR calculation are characteristic of the CCD x 512 back-illuminated L3 sensor from e2v. Dynamic range only exceeds 14-bits 1 MHz, through either amplifier

27 Technical Note Making Sense of Sensitivity It is often questioned whether or not to use EMCCD gain or whether to use EM or conventional CCD amplifiers (model dependent). The answer usually depends both on required frame rate and on light levels. Plots of Signal to Noise ratio vs Signal Intensity can be instructive in making such decisions. Here we introduce the concept of Signal to Noise in EMCCDs and discuss such plots. Part 1 - Understanding Noise Sources in EMCCDs Read Noise Read Noise in many instances can be considered the true CCD detection limit, particularly the case in fast frame rate experiments because, (a) short exposures combined with low dark current make the dark current contribution negligible and (b) faster pixel readout rates, such as 5 MHz and higher, result in significantly higher readout noise. The fundamental advantage of EMCCD technology is that gains are sufficient to effectively eliminate readout noise, therefore eliminating the detection limit. Multiplicative Noise This noise source is only present in signal amplifying technologies and is a measure of the uncertainty inherent to the signal multiplying process. For example, during each transfer of electrons from element to element along the gain register of the EMCCD, there exists only a small probability that the process of impact ionization will produce an extra electron during that step. This happens to be a small probability but when executed over more than 590 steps, a very large potential for overall EM gains result. However, the downside to this process results from the probabilities. Due to this, there is a statistical variation in the overall number of electrons generated from an initial charge packet by the gain register. This uncertainty is quantified by a parameter called Noise Factor and detailed theoretical and measured analysis has placed this Noise Factor at a value of 2 (or ~1.41) for EMCCD technology. This is an additional form of noise that must be taken into account when calculating Signal/Noise for these detectors. Note that this noise source is significantly greater from the MCP of ICCDs than from the gain register of the EMCCD. ICCDs have noise factors typically ranging from 1.5 to >2. around the mean value, in addition to the variation that already exists from the shot noise (variation from pixel to pixel or from frame to frame). Dark Current Due to the effective cooling inherent to Andor s cameras, dark current is minimized, and may often be considered practically negligible. The extent of contribution is dependent on exposure time, since the dark current is quoted in electrons/pixel/sec. It is particularly important to eliminate dark current with EMCCD technology as even single thermally generated electrons in the silicon will be amplified in the gain register just as a single photoelectron, and will appear in the final signal as a single noise spike. Fortunately, for fast frame rate experiments combined with ixon very low dark current, this noise source may be ignored. Spurious Noise Spurious Noise appears in the form of Clock Induced Charge (CIC) in EMCCD technology. CIC is independent of exposure time and generally single electron events generated during charge shift (EBI is the form of spurious noise in ICCDs and is exposure dependent). CIC is generated in every CCD but is normally buried in the readout noise. In the EMCCD however, these single electrons are amplified by the gain register just as a single photoelectron would be. In the EMCCD, CIC can in some ways be considered the true limit of detection, in that at the single photon detection level, a single photon spike will be indistinguishable from a CIC spike. Andor has specialized electronics however, that enable this source of noise to be minimized. In practical terms, ultra-weak signals of the single photon nature would be distinguishable from CIC spikes in that one could generally expect to see groupings of photon spikes from adjacent pixels, even from diffraction limited single molecule emissions. Shot Noise can be determined if the average signal is measured in electrons - by measuring in electrons, the calculation is independent of the sensor s QE i.e. the photons have already been converted to photoelectrons so the QE corrected signal is being measured. If the average signal in photons is already known (e.g. estimated from other measurements with PMTs), the shot noise can be corrected for sensor QE at that wavelength: i.e. therefore Since spurious noise is very different in nature to shot noise, it is best to consider spurious noise separately. Each EMCCD will have a measured figure for the levels of CIC spikes to be expected during a readout. This will present a figure for the average number of random spurious single electron spikes that will appear within the image. If the measured signal is at the very low photon level (one or two electrons per pixel), this noise source will be more significant. If the signal is slightly more intense than this, it may become less of an issue, and may even be filtered out. Note that often, the minimal amount of spurious noise remaining from the ixon is minor compared to the level of background photons in the image. Part 2: Signal to Noise Plots EM gain ON vs EM gain OFF (faster speeds) Figure 1 shows Signal to Noise plots derived from the specs of the back-illuminated ixon EMCCDs, read out at 10 MHz (for fastest frame rates). A photon wavelength at which the QE of the sensor is 90% is assumed. Such plots are very useful to gauge the signal intensity at which it becomes appropriate to use EM gain to increase S/N. It is clear that at 10 MHz readout, one needs to encounter relatively intense signals of > 2900 photons / pixel before it becomes advantageous to operate with EM gain off. Note that the ideal curve represents a pure Signal to Shot Noise and is shown for reference if the detector had no sources of noise, this is what the curve would appear like. Even with EM gain turned on we encounter uniformly lower signal to noise than the ideal curve. This is due to the influence of EM Multiplicative Noise, which has the effect of increasing the shot noise by a factor of 2 (~1.41). Note, Multiplicative Noise (Noise Factor) is generally higher for ICCDs. EM vs Conventional Amplifier (slower speeds) Figure 2 shows Signal to Noise plots derived from the specs of the back-illuminated ixon EMCCDs at 1 MHz (slower frame rate operation), read out either with EM gain ON or alternatively through the Conventional amplifier (i.e. standard CCD operation). Again, a photon wavelength at which the QE of the sensor is 90% is assumed. Specifically this figure applies to 897 and 888 models for which this choice of amplifier is available. Under slower speed operations, when one has the choice to read out as a conventional CCD, it can often be advantageous to do so in terms of achieving better signal to noise. Indeed the plots show that the cross-over point is at ~ 42 photons / pixel, below which it is advised still to readout through the EM amplifier with gain applied. However, one way to better understand the effects of this noise source is in terms of an addition to the shot noise of the system. Extra multiplicative noise has the same form as shot noise in that each noise type results in an increase in the variation of number of electrons that are read out of a CCD (under constant uniform illumination). Indeed multiplicative noise can be thought to contribute directly to the overall shot noise, in that one should multiply the shot noise by the noise factor when calculating overall noise. Simply put, multiplicative noise does not in any way reduce the average signal intensity or reduce the number of photons that are detected, it simply increases the degree of variation of the signal Simplified consideration of EMCCD noise From the above list it is apparent that in most uses of Andor s ixon cameras, since read noise and dark current can be virtually eliminated (i.e. the noise sources that would define the detection limit have been rendered effectively negligible), the principal sources of noise that must be considered are shot noise, noise factor (multiplicative noise) and spurious noise. It is easy to combine shot noise and multiplicative noise in an overall noise equation, using: Overall Noise = Shot Noise x 1.41 Figure 1 EM gain-on vs EM-gain-OFF signal to noise plots for backilluminated ixon EMCCDs at 10 MHz readout speed applies to 897, 860 and 888 models. Figure 2 EM gain-on vs Conventional Amplifier signal to noise plots for back-illuminated ixon EMCCDs at 1 MHz readout speed applies to 897 and 888 models

28 Technical Note ixon Ultra and ixon3 Trigger Modes Andor s EMCCDs offer a comprehensive range of internal, software and external trigger modes. Furthermore, software and internal trigger modes avail of cutting-edge firmware and SDK enhancements, delivering enhanced speed performance during complex software acquisition protocols (icam). On-head storage of multiple exposure times facilitates rapid exposure time switching upon receiving a trigger. Part 1 - External Trigger Modes The ixon back-illuminated range of cameras have several different external trigger modes: External Trigger in Frame Transfer Mode (Simultaneous Exposure and Readout) In this mode, the camera sits in its External Keep Clean cycle, which can be interrupted by the external trigger with a jitter of only a couple of microseconds (exact time dependent on camera model). Upon receiving a trigger, the camera stops all vertical clocking and waits for the programmed user delay period before starting the read phase. During the readout phase the Image area is transferred rapidly to the Storage area. The Storage area is then read out in the normal way. similar to the external trigger mode discussed above, in that the type of keep cleans are identical and the exposure is started by the positive edge of the trigger pulse. Where these two trigger modes differ is in how the exposure time is controlled. With standard External Trigger, the user (via software) controls the exposure time. With External Level Trigger mode, the level of the trigger pulse (i.e. the time spent in the on state of the TTL signal) controls the exposure time. The exposure period starts on the positive edge of the trigger pulse and stops on the negative edge. The exposure is physically ended by shifting the Image area into the Storage area. The Storage area is then read out in the normal manner. On completion of the readout, the External Keep Clean cycle is started again. Part 2 icam Fast Exposure Switching icam delivers: Enhanced software triggering during acquisition highly efficient upload of acquisition parameters from software to camera with minimized overheads. Ring Mode ultimate in exposure switching during multi-channel protocols. Software pre-loads up to 16 acquisition channels onto the camera. Enhanced external trigger mode optimized speed performance across all of Andor s comprehensive external trigger options. Asynchronous frame transfer mode (AFTM) enhanced speed and synchronization in overlapped/frame transfer acquisition mode. Bi-directional communication between PC and camera. Further enhanced baseline stability takes Andor s marketleading quantitative baseline stability to a whole new level. Third party software compatibility most popular imaging suites can take advantage of icam. Enhanced data exchange between camera and PC icam allows for faster frame rates in software by using dedicated timing patterns that shorten unnecessary overhead times. These time lags prevent other EMCCD and interline cameras on the market from achieving fast frame rate during complex experimental protocols. Furthermore, icam s Ring Mode offers the capacity to use up to 16 different timing patterns uploaded into the camera head, thus external triggers can result in virtually instantaneous switching between channels, facilitating unparalleled synchronization with other peripheral equipment such as filter wheel, laser-aotf or z-stage. Once the read out is complete the camera continues to wait for the next external trigger event. While the camera is waiting for the trigger event the shift register is continually clocked but the Image and Storage areas are not. On the next trigger the camera again waits for the programmed delay before starting the read out phase. The camera continues in this cycle until the number of images requested have been captured. Since the Image area is not cleaned between trigger events, the exposure time is defined by the time between trigger events. This sequence is illustrated in Figure 1. External Trigger in Non-Frame Transfer Mode In this mode, the camera is once again sitting in External Keep Clean. As can be seen from Figure 2, the External Keep Clean cycle runs continuously until the external trigger event is detected, at which point the exposure phase starts. Once the exposure time has elapsed the charge built up in the Image area is quickly transferred into the Storage area. From the Storage area the charge is read out as normal. At the completion of the read out the camera restarts the External Keep Clean cycle and will perform a minimum number of cleans before accepting the subsequent trigger event. Fast External Trigger in Non-Frame Transfer Mode This mode is for the most part identical to External Trigger Mode and differs in only one respect: after an acquisition and readout, the camera will not wait for a sufficient number of keep clean cycles to have completed before allowing a trigger event to start the next acquisition. As a result Fast External Trigger allows a higher frame rate than standard external trigger mode. Fast External Trigger is most useful in those cases where there is very little light falling on the sensor outside of the exposure times and the user is looking for fast frame rates. External Exposure in Non-Frame Transfer Mode Figure 3 shows a timing scheme for External Exposure mode, another external trigger option for non-frame transfer readout. This mode is icam encompasses a set of unique innovations that empower Andor ixon cameras to operate with complete acquisition efficiency through Andor iq multi-dimensional microscopy suite and other third party imaging software packages. ixon3 897 with icam 23 Other EMCCD 12 Frame Rate (fps) Table 1 - Comparison of exposure switching speed in ixon3 897 vs. competing EMCCD camera with same 10 MHz pixel readout. Freerun imaging sequence involving dual channel acquisition protocol using rapid toggle between 1 ms and 2 ms for each channel. Both cameras have the same 512 x 512 back-illuminated sensor. Imaging cameras in a heightened state of readiness Andor s icam technology is a combined firmware and software innovation, a highly efficient and performance-optimized solution that is now integrated across all new Andor imaging cameras and Andor s iq and SDK software platforms. icam functionality has been integrated into a number of popular third party software drivers, including MDC Metamorph and Nikon NIS Elements. icam offers heightened EMCCD performance during tightly synchronized and complex multi-dimensional microscopy experiments. Using state of the art bi-directional communication between camera and PC, icam is particularly effective for multi-channel acquisitions during which different exposure times are rapidly toggled between channels, whether software triggered or hardware (externally) triggered, with absolute minimal overheads, as demonstrated in the data in Table 1. Figure 1: External Trigger in Frame Transfer mode Figure 2: External Trigger in non-frame Transfer mode Figure 3: External Exposure in non-frame Transfer mode 54 55

29 Technical Note ixon Ultra Camera Link Output for Direct Data Access Under standard operation the ixon Ultra uses the USB interface for all control and data transfer with the PC. However, some users require a more direct access to the image data stream, in order that they can perform real-time analysis, possibility using external hardware. Such operation can be particularly important for rapid closed feedback applications such as adaptive optics. Direct real time access to data can also be useful for data intensive applications such as superresolution microscopy or whole genome sequencing, whereby it can be desirable to carry out real time processing of data on an external GPU for example. In order to facilitate such functionality, the ixon Ultra also includes a Camera Link output. The Camera Link output conforms to the specification as defined by the Automated Imaging Association (AIA). Note that the Camera Link on the ixon Ultra is an output only i.e. the serial communication interface used for camera control has NOT been implemented. When the Camera Link output is in use all camera control is still transmitted via the USB interface. The image data stream is still transmitted over the USB interface, which allows other more basic analysis to be run on the PC to monitor other characteristics such as the signal level or visual feedback. The Camera Link output is a base configuration (3-tap interface) running at 40 MHz. All data over the Camera Link interface is 16-bit greyscale. The exact pixel data type is Little Endian 16-bit unsigned integer. The Camera Link channel intercepts the image data stream in the camera head immediately after the on-head FPGA processing step, but before the USB frame buffer, it therefore undergoes the same amount of on-head image processing. The data stream only contains pixel data for those pixels within the camera user defined region of interest (ROI) and hence the length of LVAL and FVAL (Camera Link LINE and FRAME, respectively) will be dependent on the ROI defined. The user could set the camera to always transmit the same ROI and then use their Camera Link frame grabber card to extract a different ROI but they will not see any speed increase normally associated with smaller ROIs. Users must ensure the data stream is compatible with any limitations of your particular Camera Link card. For example, when acquiring a sub-region ensure the number of bytes per line and per frame match the granularity expected by the card. Latency Test The Camera Link channel intercepts the image data stream in the camera head immediately after the on-head FPGA processing step, but before the USB frame buffer, hence minimizing any latency or jitter. Figure 1 is an oscilloscope trace showing the camera FIRE output (which goes high during the exposure), the Camera Link FRAME (FVAL) and the Camera Link LINE (LVAL). As you can see from the figure the delay between the exposure completing and the data appearing on the Camera Link output is ~500 µs, and most of this time is accounted for by the time it takes to move the image from the Image Area to the Storage Area on the frame transfer sensor (268 µs). Image Integrity test The Camera Link output has been tested using the Neon Camera Link card from Bitflow. Figure 2 show screen captures from Solis and CiView (from Bitflow), with full resolution and a 128 x 128 subregion. It is clear that the USB 2.0 and Camera Link interface each show identical images. The ixon Ultra Camera Link output should be compatible with any Camera Link interface card, but only Bitflow has been tested in house. Figure 1: Latency Vertical shift rate = 0.5µs, Time from falling edge of FIRE to rising edge of Camera Link FRAME = 494.4µs (268µs of this time is the time required to shift the image into the mask area of the sensor). Ch1: Camera FIRE (Yellow); Ch2: CL FVAL (Blue); Ch3: CL LVAL(Pink) (a) Operation Camera is configurable via the USB 2.0 (897 model) or USB 3.0 (888 model) interface as normal through acquisition software (e.g. Andor Solis) or SDK. As the Camera Link output is completely independent of the camera control interface, all the features of the camera are fully available. To use the Camera Link interface simply enable the Camera Link output. In SDK, or in AndorBasic in Solis, issue the command SetCameraLinkMode(mode) where mode=0 disables and mode=1 enables the Camera Link output. Images must be requested over the USB to prevent the USB buffer from overflowing. Note, any jitter associated with requesting data over the USB will not introduce any jitter on the Camera Link interface provided the USB buffer does not overflow. (b) Figure 2 - Images captured using Solis (left) via USB and simultaneously in CiView (right) via Camera Link interfaces. (a) Full resolution (b) 128 x 128 ROI 56 57

30 Notes... The EMCCDs made by Andor are perfect cameras for our superresolution microscopy on biological materials. The EMCCDs have incredibly high quantum yields and extremely low noise. You can t ask too much more out of the cameras! Prof. Paul Selvin, Department of Physics, University of Illinois, USA QE curves of ixon models vs emission of common fluorophores Front cover image: Comparison of conventional and STORM images of mitochondria in a mammalian cell. (left panel) Conventional image of the left part of the cell. (middle panel) 3D STORM image of the middle part of the cell. (right panel). The xy crosssection of the STORM image of the right part of the cell. The relative sensor size of the ixon models Courtesy of Xiaowei Zhuang, Harvard University, USA 58 59

31 Customer Support Andor products are regularly used in critical applications and we can provide a variety of customer support services to maximize the return on your investment and ensure that your product continues to operate at its optimum performance. Andor has customer support teams located across North America, Asia and Europe, allowing us to provide local technical assistance and advice. Requests for support can be made at any time by contacting our technical support team at andor.com/support. Andor offers a variety of support under the following format: On-site product specialists can assist you with the installation and commissioning of your chosen product Training services can be provided on-site or remotely via the Internet A testing service to confirm the integrity and optimize the performance of existing equipment in the field is also available on request. A range of extended warranty packages are available for Andor products giving you the flexibility to choose one appropriate for your needs. These warranties allow you to obtain additional levels of service and include both on-site and remote support options, and may be purchased on a multi-year basis allowing users to fix their support costs over the operating life cycle of the products. Head Office 7 Millennium Way Springvale Business Park Belfast BT12 7AL Northern Ireland Tel: +44 (0) Fax: +44 (0) North America 425 Sullivan Avenue Suite 3 South Windsor, CT USA Tel: Fax: Japan 4F TK Sarugakucho Building Sarugaku-Cho Chiyoda-Ku Tokyo Japan Tel: +81 (0) Fax: +81 (0) China Room 1213, Building B Luo Ke Time Square No. 103 Huizhongli Chaoyang District Beijing China Tel: +86 (0) Fax: +86 (0) Find us on Andor, the Andor logo and ixon are trademarks of Andor Technology Ltd. Changes are periodically made to the product and specifications are subject to change without notice. LiiXB 1214